Analysis Of The Diagnostic Efficacy Of Contrast-enhanced Ultrasound LI-RADS M Criteria Modified Version In Non-HCC Malignancies

Objective:By analyzing and comparing the diagnostic efficacy of contrast-enhanced ultrasound(CEUS)LI-RADS M criteria and its modified version,to investigate its diagnostic value in non-HCC malignancies.Methods: From January 2012 to October 2022,346 patients with focal liver lesions were examined by conventional ultrasound and CEUS in the Department of Ultrasound,Shengjing Hospital Of China Medical University.The lesions were classified according to the CEUS LI-RADS 2017 version,the following three criteria are used as the diagnostic criteria of M category: criterion 1(original CEUS LI-RADS M criteria): Rim APHE,early washout(< 60s),marked washout;criterion 2(modified version):Rim APHE,early washout(< 45s),marked washout;criterion 3(modified version):Lesions from patients with normal CA19-9 with elevated AFP were excluded on the basis of meeting criterion 1.Lesions excluded by criterion 2 and 3 were reclassified into 5 categories.The sensitivity,specificity,PPV,NPV and accuracy of each creteria diagnosis were calculated,and the diagnostic value was judged and compared.Results:A total of 401 lesions were identified in 346 patients,which were pathologically confirmed,including 343 HCC lesions,36 non-HCC liver malignancy lesions,and 22 benign lesions.According to the CEUS LI-RADS classification,107(26.7%)lesions were classified as category M.Among LR-M lesions,the number of HCC lesions was 71(66.4%),and most of them were moderately and poorly differentiated HCC.The sensitivity,specificity,accuracy,PPV and NPV of criterion 1 for diagnose non-HCC malignancy were94.4%、80.0%、81.3%、31.8% and 99.3%,respectively.The specificity and accuracy of criterion 2 and 3 in the diagnosis of non-HCC malignancies was higher than that of criterion 1(91.5%vs.80.0% 、 88.5%vs.80.0%;90.0%vs.81.3%、87.8%vs.81.3%;all P<0.05).Notably,the sensitivity of criterion 2 was lower than criterion 1(75.0%vs94.4%;P<0.05),but the sensitivity between criteria 1and 3 was not statistically significant differences(94.4%vs.80.6%;P> 0.05).The PPV of criterion 2 and 3 were 46.6% and 40.8%,which were higher than criterion 1,but the difference was not statistically significant(P> 0.05).The NPV of criterion 2 and 3 were97.4% and 97.9%,which were lower than criterion 1,but the difference was not statistically significant(P> 0.05).The sensitivity,specificity,accuracy,PPV,and NPV of the LR-5 categories for HCC were 66.2%,91.4%,69.8%,97.8%,and 31.4%,respectively.After excluding by criterion 2,3,the sensitivity of diagnosis increased compared with the original LR-5 category,statistically significant difference(all P <0.05).After reclassification of criterion 2,the specificity of diagnosing HCC decreased,and the difference was statistically significant(P <0.05).There was no significant difference in the specificity of diagnostic HCC between criterion 3 reclassification and the original LR-5 category(P> 0.05).After reclassification by criterion 2 and 3,PPV decreased and NPV increased,but the difference was not statistically significant(P> 0.05).Conclusion: CEUS LI-RADS M has high sensitivity for non-HCC liver malignancy,but low specificity due to a large number of HCCs in M category.Adjusting early washout time and the combination of tumor markers can improve the specificity of diagnosis.If the washout time is advanced,the sensitivity of diagnosis will reduce,and the selection of washout time still needs to be studied.Combining with tumor markers can improve the specificity while maintaining the sensitivity.It is a potential method for the diagnosis of non-HCC liver malignant tumors.