| Section 1 Quantitative assessment of cholestasis liver fibrosis with multifrequency magnetic resonance elastography:a preliminary experimental studyObjective:1.To evaluate the diagnostic performance of magnetic resonance elastography(MRE)and the optimal cutoff values for staging cholestasis liver fibrosis and to determine the correlation between the stiffness value and the stage of hepatic fibrosis.2.To assess the optimal frequency in quantitatively assessing liver stiffness in rabbits.Materials and Methods:1.The different stages(F0-F4)of cholestasis liver fibrosis were induced by common bile duct partial ligation(CBDPL)in experimental rabbit.MRE was performed using vibration frequencies of 60Hz,80Hz and 100Hz to assess liver stiffness values of normal contrast rabbits and experimental rabbits after CBDPL.2.Assessment of liver stiffness value in rabbits:Two observers "blinded" to the treatment given the rabbits measured the stiffness value of rabbit livers.Regions of interest(ROIs)excluding the liver margins,major blood vessels and artifacts were drawn on selected three slices.Three ROIs per slice were placed and averaged to produce a measure of stiffness value.3.Statistical analysis:A single-sample Konogorov-Smimov test was performed on the stiffness values measured at different stages of hepatic fibrosis.The difference in stiffness between different stages was compared using the one-way analysis of variance(LSD)method if the result fitted normal distribution.Statistical analysis results were considered to be significant at P<0.05.Correlation between the stage of hepatic fibrosis and the liver stiffness values were analyzed with Pearson 's correlation test.The correlation coefficient R was calculated and a scatter plot was drawn to obtain R2.The optimal cutoff value and efficiency of liver stiffness measurement for the determination of fibrosis stages were determined by a receiver operating characteristics(ROC)curve analysis.The area under the curve(AUC)was measured and the best frequency is initially determined.Analyses were performed by using software(SPSS software for Windows,22.0,Chicago).Results:1.Statistical differences between different stages of fibrosis:There were significant differences in the stiffness values between different stages of fibrosis except F0 and F1 at the frequency of 60Hz,F2 and F3 at the frequency of 60Hz and 80Hz,F3 and F4 at the frequency of 80Hz and 100Hz.2.Correlation analysis and diagnostic performance evaluation:There was a significant positive correlation between different stages of fibrosis and liver stiffness values at the frequency of 60Hz,80Hz and 100Hz with the R value of 0.675,0.732 and 0.745 respectively(P<0.001).Areas under ROC curve was 0.89 for fibrosis stage F?2 at the frequency of 60Hz and the others were all greater than 0.91.Especially at the frequency of 80Hz,the areas under ROC curve were all greater than 0.95.At the frequency of 60Hz,optimal stiffness cutoff values of 2.65,2.88,3.07 and 3.33kPa showed ?F1,?F2,?F3 and F4,respectively.At the frequency of 80Hz,optimal stiffness cutoff values of 3.34,3.81,3.96 and 4.11kPa showed ?F1,?F2,?F3 and F4,respectively.At the frequency of 100Hz,optimal stiffness cutoff values of 4.69,4.99,5.31 and 5.55kPa showed ?F1,?F2,?F3 and F4,respectively.Conclusion:1.MRE showed promising capability of differential diagnosis different stages of cholestasis liver fibrosis.The diagnostic performance were all high in different stages of fibrosis at the frequency of 60Hz,80Hz and 100Hz.There was a significant positive correlation between different stages of fibrosis and liver stiffness values.2.The optimal frequency in quantitatively assessing liver stiffness in rabbits was 80Hz.Section 2 Quantitative evaluation of periductal benign and malignant lesions with magnetic resonance elastography:a case-control studyObjective:To explore the value of MRE in quantitatively assessing periductal benign and malignant lesions,determining the optimal cutoff value and evaluating the diagnostic performance.Materials and Methods:The contrast-enhanced multiple-phase(CEMP)computed tomography(CT)and magnetic resonance imaging(MRI)findings,CA199 and elasticity values of 60 patients with periductal benign or malignant lesions proved by surgery and pathology in our hospital from January 2014 to December 2016 were collected.And CEMP CT,MR and liver MRE scans were performed in all these patients.There were 44 patients admitted eventually.Among them,there were 16 cases with periductal benign lesions including hepatolithiasis,cholangitis and fibrosis(male-to-female ratio,5:11;age range,21-82 years;mean age,54.9 years)and 28 cases with periductal malignant lesions(cholangiocarcinoma)(male-to-female ratio,21:7;age range,31-81 years;mean age,57.2 years).CEMP CT?MRI findings,liver elasticity values and CA199 of the two groups were analyzed.Liver elasticity values were compared between the following groups:periductal benign lesions and relatively normal region of liver parenchyma 1,periductal malignant lesions and relatively normal region of liver parenchyma 2,periductal benign and malignant lesions,relatively normal region of liver parenchyma 1 and 2.Continuous variables were tested by the t-test or paired t test and categorical variables were tested by the ?2 test or continuity correction ?2 test.P values less than 0.05 were considered statistically significant.Correlation between the elasticity values of the periductal malignant lesions and the relative normal region of liver parenchyma 2 were analyzed with Pearson's correlation test.The correlation coefficient R was calculated.A ROC curve was performed to determine the cut-off values of CA199 and elasticity value differentiating periductal benign lesions from malignant lesions,and ROC curves of CEMP CT?MRI findings and the three combined methods differentiating periductal benign lesions from malignant lesions were also performed.Diagnostic efficacy and the AUC were compared among the four diagnostic methods.Result:When compared by pairs,the elasticity values of the periductal benign and malignant lesions and the relative normal region 1 and 2 were all statistical differences with P values of<0.001,<0.001,<0.001 and 0.006 respectively.Periductal malignant lesions had significantly greater mean elasticity value than periductal benign lesions(10.1 kPa vs 5.9 kPa,P<0.001).But there was no significant correlation between the elasticity values of the periductal malignant lesions and the relative normal region(R=0.166,P=0.399).There were statistical differences in CEMP CT/MRI findings(P<0.001),CA199 values and the three combined methods(P<0.001)differentiating periductal benign and malignant lesions(P=0.008).The optimal cutoff value of elasticity value to differentiate periductal benign lesions from malignant lesions was 6.42 kPa with a sensitivity of 92.9%,specificity of 82.1%,accuracy of 88.1%and AUC of 0.934.And the optimal cutoff value of CA199 value was 140.6 U/ml with a sensitivity of 71.4%,specificity of 85.7%,accuracy of 78.6%and AUC of 0.783.The sensitivity,specificity,accuracy and AUC of CT/MRI findings differentiating periductal benign lesions from malignant lesions were 78.6%,81.2%,79.5%and 0.799 respectively.The sensitivity,specificity,accuracy and AUC of the three combined methods differentiating periductal benign lesions from malignant lesions were 53.6%,100%,70.5%and 0.768 respectively.Conclusion:MRE showed promising capability of differential diagnosis of periductal benign and malignant lesions.The optimal cutoff value was 6.42 Kpa and the diagnostic accuracy was higher than that of CA199 and CEMP CT/MRI findings.Section 3 Quantitative Evaluation with MRE and CEMP imaging diagnosis of IMCC and HCC with cirrhosis:a comparative studyObjective:1.To explore the value of MRE in quantitative evaluation of intrahepatic mass-forming cholangiocarcinoma(IMCC)and hepatocellular carcinoma(HCC)in cirrhotic patients,determine the optimal cutoff value and evaluate the diagnostic performance.2.To evaluate the CEMP CT and MRI features of IMCC and HCC in patients with cirrhosis and determine whether specific findings suggest correct diagnosis of IMCC.Materials and Methods:A retrospective case-control study was conducted.50 IMCC patients and 50 HCC controls with underlying liver cirrhosis consecutively registered in our hospital between January 2010 and December 2016 were analyzed.All patients were proved by surgery and pathologically.Among them,liver MRE scans were performed on 28 IMCC patients and 28 HCC controls,and CEMP CT and/or MRI scans were performed on other 22 IMCC patients and 22 HCC controls.Liver stiffness was compared between tumours and non-tumor areas in patients with liver MRE scan.Meanwhile,the tumor markers(CA199?AFP and CEA)of these patients were also evaluated.Signal and/or density on each phase and enhancement patterns of the tumors and accompanying findings were all evaluated in patients with CEMP CT and/or MRI scans.Continuous variables were compared with the student's t test.Categorical variables were tested using ?2 test or Fisher's exact test.A conventional p value less than 0.05 was considered statistically significant.A ROC analysis was performed to determine the cut-off values of stiffness value?AFP and CA199 differentiating IMCC from HCC with cirrhosis.Sensitivity,specificity,accuracy and the AUC were compared.Results:1.When compared by pairs,the stiffness values of the 28 IMCC and non-tumor area 1,28 HCC and non-tumor area 2,28 IMCC and 28 HCC in cirrhotic patients were statistically different with P values of<0.001,<0.001 and 0.003.There was also statistically different in CA199(P=0.022)and AFP(P=0.013)values of IMCC and HCC in cirrhotic patients.The optimal cutoff value of stiffness value to differentiate IMCC from HCC with cirrhosis was 7.24 kPa with a sensitivity of 71.4%,specificity of 67.9%,accuracy of 68.2%and AUC of 0.702.And the optimal cutoff value of CA199 value was 19.3 U/ml with a sensitivity of 66.7%,specificity of 68.2%,accuracy of 68.2%and AUC of 0.703.The optimal cutoff value of AFP value was 37.45 ng/ml with a sensitivity of 68.2%,specificity of 71.4%,accuracy of 70.5%and AUC of 0.722.When combined MRE with CA199,the sensitivity was 60.7%,the specificity was 85.7%,and the accuracy was 73.2%.When combined MRE with AFP,the sensitivity was 28.6%,the specificity was 96.4%,and the accuracy was 62.5%.2.Enhancement pattern,signal on MRI delayed phase(P<0.001),as well as accompanying characteristics of maximum diameter(P=0.010),capsule retraction(P=0.001),portal vein invasion(P=0.002),bile duct dilation(P=0.007)and abdominal lymphadenectasis(P<0.001)were all different between 22 IMCC and 22 HCC in cirrhotic patients.On CEMP CT and MRI,the most common enhancement patterns of IMCC were progressive enhancement pattern(P=0.001 or P<0.001)and rim-like enhancement pattern(CT,P=0.021).While washout pattern was the most frequent enhancement pattern of HCC(P<0.001).12/15(80%)IMCC lesions while only 2/21(9.5%)HCC lesions with cirrhosis appearing as hyperintense at delayed phases on CEMP MRI(P<0.001).Conclusion:1.MRE showed promising capability of differential diagnosis between IMCC and HCC in cirrhotic patients.The optimal cutoff value was 7.24 kPa and the diagnostic accuracy was similar to CA199 and AFP.When MRE combined with CA199,the diagnostic accuracy was the highest.2.It should be seriously considered the diagnosis of IMCC in cirrhotic patients that progressive and/or peripheral rim-like enhancement on CEMP CT and MRI,hyperintense at delayed phase on CEMP MRI with capsule retraction,portal vein invasion,bile duct dilation,abdominal lymphadenectasis,and/or CA199 positive. |
PDF Full Text Request