| ObjectiveTo evaluate liver function according to the degree of biliary system visualization on Gd-BOPTA-enhanced MRI in hepatic cirrhosis patients.Materials and MethodsData from 523 patients with suspected liver disease,who underwent hepatic contrast enhanced MR imaging with Gd-BOPTA in the authors'hospital between April 2018 and May 2019,were included.Patients who had undergone surgery involving the biliary tract(n=54),those diagnosed with primary sclerosing cholangitis(n=29),cholestasis(n=16),or primary biliary diseases(n=12)were excluded.Due to missing biochemical parameter data(n=48)or image quality issues(n=20),68 additional patients could not be enrolled.Ultimately,343 patients were enrolled in the present study,including 132 patients with normal liver function and no underlying chronic liver disease(NLF)and 211 patients with hepatic cirrhosis(hepatic cirrhosis group).The causes of hepatic cirrhosis in these 211 patients include hepatitis B(n=154),hepatitis C(n=28),alcoholic hepatitis(n=24),autoimmune hepatitis(n=3),and cryptogenic cirrhosis(n=2).Patients in the hepatic cirrhosis group were sub-classified into one of the following three groups,according to Child-Pugh classification:Child-Pugh class A(LCA,n=107);Child-Pugh class B(LCB,n=70);and Child-Pugh class C(LCC,n=34).This study was approved by our hospital ethics committee.Written informed consent for participation was obtained from all patients.Biochemical parameter data for total bilirubin,serum albumin,and PT within 2 weeks before or after liver MRI examination,were collected.Hepatic cirrhosis was verified based on radiological features extracted from ultrasonography,computed tomography,or MRI(surface nodularity,relative enlargement of the left lobe,marginal dullness,ascites,splenomegaly and the formation of varices)in all patients.MRI was performed using a 3 Tesla system(MAGNETOM Prisma,Siemens,Germany)equipped with a phased-array body coil.For each patient,MR images were independently evaluated by two experienced academic radiologists,who were blinded to all clinical and radiological data,and assessed all MR images at the pre-contrast and hepatobiliary phases.To quantitatively measure the relative signal intensity(SI)of biliary contrast enhancement at the hepatobiliary phase,regions of interest(ROIs)were placed in an optimal area of the upper end of the common bile duct(SI up)and lower end of the common bile duct(SI low)in the axial image.The ROI area was 4-10 mm2.To reduce error,the SI of the erector spinae muscle(SI m)was measured using the same size ROI in the erector spinae muscles.Finally,the relative enhancement of the biliary system(RE)was calculated as follows:RE=(SI up/SI m+SI low/SI m)/2.To quantitatively measure SI of the liver parenchyma(SI.of liver),the ROIs were placed at 2 locations in the left lobe(lateral and medial segments)and 2 locations in the right lobe(anterior and posterior segments),and the SI of the erector spinae muscle(SI m)was measured using the same size ROI in the erector spinae muscles.The average value of the four SIs of the liver parenchyma and SI m were used for data analysis.The liver-to-muscle ratio(LMR)was calculated as follows:LMR=SI of liver/SI m.These images were acquired using the same window width(800)and window level(400).All data were analyzed using SPSS version 22.0(IBM Corporation,Armonk,NY,USA)and MedCalc 12(Medcalc Software,Mariakerke,Belgium).The RE of the biliary system and serum parameters(total bilirubin,PT,and albumin)were the continuous variables.Pearson's correlation analysis,one-way analysis of variance(ANOVA),and post hoc pairwise comparisons were performed for statistical analysis;differences with P<0.05 were considered to be statistically significant.Diagnostic performance in quantitatively evaluating liver function was assessed using receiver operating characteristic(ROC)curve analysis.The corresponding optimal cut-off values and areas under the ROC curve were determined.Results1.RE of the biliary system were negatively correlated with total bilirubin(r=-0.542,P<0.01)and PT(r=-0.520,P<0.01).RE of the biliary system was positively associated with albumin level(r=0.555,P<0.01);2.The RE of the biliary system were positively associated with LMR in all patients(r=0.457,P<0.001);3.The RE of the biliary system in the NLF group were significantly higher than those of the LCA group(3.79 ± 0.69 versus 3.01± 0.87;P<0.001),LCB group(3.79± 0.69 versus 1.73± 0.87;P<0.001),and LCC group(3.79±0.69 versus 0.58± 0.07;P<0.001).The RE of the biliary system in the LCA group was significantly higher than in the LCB(3.01 ± 0.87 versus 1.73 ± 0.87;P<0.001)and LCC(3.01 ± 0.87 versus 0.58 ± 0.07;P<0.001)groups.There were significant differences between the LCB and LCC groups(P<0.001)with regard to RE of the biliary system;4.RE of the biliary system demonstrated significantly better diagnostic performance in distinguishing patients with normal liver function and no underlying chronic liver from those with Child-Pugh A,patients with Child-Pugh A disease from those with Child-Pugh B disease,and patients with Child-Pugh B disease from those with Child-Pugh C disease than LMR(AUC,0.751[95%confidence interval:0.691,0.805]vs 0.571[95%confidence interval:0.505,0.634],P<0.0001,AUC,0.823[95%confidence interval:0.758,0.876]vs 0.695[95%confidence interval:0.621,0.762],P=0.002,AUC,0.854[95%confidence interval:0.771,0.915]vs 0.642[95%confidence interval:0.542,0.734],P=0.001).Analysis of the ROC curves yielded a cut-off value of ?3.001 for RE of the biliary system(sensitivity,0.90;specificity,0.83)for patients in the NLF group.A cut-off value of?2.301 for RE of the biliary system(sensitivity,0.82;specificity,0.85)was found for patients in the LCA group.ConclusionThe degree of biliary system visualization of Gd-BOPTA-enhanced MRI may be used as an imaging-based,quantifiable metric to estimate liver function. |
PDF Full Text Request