Standardizing Magnetic Resonance Portovenography And Its Application In Chronic Liver Disease: An Experimental Study

Objective: To standardize the technique of magnetic resonance (MR)portal venography in healthy piglets by utilizing digital subtractionangiography (DSA) as the gold standard.Materials and Methods: Sixteen healthy piglets were enrolled into thisstudy and underwent MR perfusion scans in the condition of generalanesthesia, and the dynamic scanning volume encompassed the main portalvein. By the GE perfusion commercial software, the time-signal curve ofportal vein was generated automatically to obtain the time of portal vein peakenhancement for optimizing imaging time of MR portal venography.Twenty-four hours after the dynamic study, the fat-suppressedthree-dimensional fast gradient echo sequence (3D FSPGR) MR coronalenhancement scanning was carried out in the experimental animals accordingto the above-mentioned optimal imaging time of MR portal venography, andthe coronal imaging data were used for multiplanar volume reconstruction todisplay the main portal vein and principal branches and measure their caliber.Subsequently, each piglet underwent indirect portography via superiormesenteric artery and splenic artery to demonstrate the mesenteric veins,splenic vein and main portal vein, and to measure their caliber. The data ofindirect portography via superior mesenteric artery were randomly selectedfor calculating the best portal vein imaging time by this angiography. The optimal imaging time of portovenography, or the diameters of the main portalvein (MPV) and branches between MR and DSA were statistically analyzed.Results: (1) The optimal imaging time of MR portal venography wassignificantly longer than that of DSA venography (31.61±2.03 s vs. 14.40±0.75 s, respectively, p < 0.05), but there was a good correlation between MRand DSA (r = 0.749, p < 0.05). (2) The diameters of MPV, superior mesentericvein and splenic vein on MR portal venography were significantly larger thanthose on DSA venography (10.67±1.59 mm vs. 7.65±1.17 mm, 8.87±1.43mm vs. 5.74±1.05 mm, and 7.77±1.18 mm vs. 5.03±0.98 mm, respectively,all p < 0.05), but there were significant correlations (r = 0.913, 0.840, 0.834,respectively, all p < 0.05).Conclusion: The standard MR portal venography can be helpful foraccurately evaluating the portal system, which might be the model for thefurther study on developing the standard MR portal venography techniqueused in experimental portal hypertension secondary to liver cirrhosis. Objective: (1) To evaluate the variation of time to peak (TTP) ofportal-vein enhancement of standard MR portovenography by previous MRperfusion scans in adult piglets with experimental chronic liver disease duringthe course of modeling the disease; (2) To evaluate the variation of diametersof main portal vein and principal branches measured on standardizingmagnetic resonance portovenography in adult piglets with chronic liverdisease during the course of modeling the disease; (3) To determine the MRreference standards of TTP of portal vein and calibers of the main portal veinas well as its principal branches shown on standard MR portal venography inthe piglets with hepatic fibrosis and cirrhosis confirmed by histologic analysisby receiver operating characteristic curve analysis, and to improve thediagnostic efficacy of the experimental chronic liver disease during the courseof modeling the disease on the basis of the correlations of with TTP of portalvein and calibers of portal vein system with the laboratory liver reservefunction.Materials and Methods: 16 healthy adult piglets were injected 40%carbon tetrachloride (CCL4) oil solution at 0.25 ml per kilo weight intoperitoneal cavity twice a week until 16th week after the beginning ofmodeling the disease, and received 5% alcohol as the drinking water to establish of liver cirrhosis models. After 16th weekend, all the animalsunderwent oral administration of CCL4oil solution at 0.75 ml per kilo weightfor further modeling the disease because of peritoneal adhesion until 21thweekend. These piglets underwent MR examination and hepatic functionlaboratory tests on 0, 5th, 9th, 16th and 21th weekend after the beginning ofmodeling the disease. One-way ANOVA test was used to compare thedifference in TTP of portal vein, in their calibers of the main portal vein aswell as principal branches, or in hepatic function markers among the pigletson different weekends after the beginning of modeling the disease, and LSDor Dunnett T3 method was used to perform the comparisons between thepiglets on different weekends. Receiver operating characteristic analysis wasused to assess the diagnostic efficacy of TTP of portal vein, and those of theircalibers of the main portal vein and principal branches for hepatic fibrosis andcirrhosis, and to evaluate the variating the patterns of calibers of main portalvein (MPV), superior mesenteric vein (SMV) and splenic vein (SPV) byusing curve estimation. The Pearson correlation analysis was used to study therelevance between TTP of portal vein or the diameters of splenic vein, andhepatic function markers on different weekend or at different statusesincluding hepatic fibrosis and cirrhosis.Results: The mean values of TTP of portal vein obtained by MRperfusion scans on 0, 5th, 9th, 16th and 21th weekend after the beginning ofmodeling the disease were 31.81±2.03 s, 35.95±1.62 s, 41.36±2.44 s,41.70±2.74 s and 42.44±3.10 s, respectively; and the values at the statusesof hepatic normal, fibrosis and cirrhosis were 31.81±2.03 s, 39.54±3.49 sand 42.44±3.10 s, respectively. Statistical analysis showed that the TTP ofportal vein increased progressively during the course of modeling the disease, and those were significant difference among the piglets on different weekendsor at different statuses of the disease (all p < 0.05). On 0, 5th, 9th, 16th and21th weekend after the beginning of modeling the disease, the diameters ofMPV were 8.82±0.78 mm, 9.05±0.58 mm, 9.46±0.56 mm, 9.61±0.53mm and 9.80±0.62 mm, respectively; SMV were 7.04±0.85 mm, 7.08±0.78 mm, 7.79±0.55 mm, 8.10±0.55 mm and 7.84±0.53 mm, respectively;and the diameters of SPV were 5.83±0.61 mm, 6.32±0.58 mm, 6.65±0.46mm, 7.12±0.61 mm and 7.37±0.50 mm respectively. At the statuses ofhepatic normal, fibrosis and cirrhosis, MPV were 8.82±0.78 mm, 9.36±0.59mm and 9.80±0.62 mm, respectively; the diameters of SMV were 7.04±0.85 mm, 7.63±0.75 mm and 7.84±0.53 mm respectively; and the diametersof SPV were 5.83±0.61 mm, 6.67±0.63 mm and 7.37±0.50 mm,respectively. The calibers of the main portal vein and principal branchesincreased during the course of modeling the disease, and that there weresignificant difference among the piglet on different weekends or at differentstatuses of the disease (all p < 0.05). For the diagnosis of hepatic fibrosis, theareas under the ROC curves of TTP, MPV, SMV and SPV were 0.981, 0.696,0.711 and 0.848, respectively; and TTP and SPV could be the better index forthe diagnosis (all p < 0.05). For the diagnosis of liver cirrhosis, the areasunder the ROC curves of TTP, MPV, SMV and SPV were 0.740, 0.697, 0.563and 0.823, respectively; and TTP and SPV could be the better index for thediagnosis (all p < 0.05). At 0, 5th, 9th, 16th and 21th weekend after thebeginning of modeling the disease, the mean values of seralbumin were 30.11±3.46 g/L, 28.15±3.23 g/L , 25.84±4.48 g/L, 25.87±2.27 g/L and 23.90±4.26 g/L, respectively; the mean values of serum bilirubin were 1.39±1.45pmol/L, 8.75±5.00 pmol/L, 6.72±3.30 pmol/L, 4.76±1.96 pmol/L and 13.39±14.51 pmol/L, respectively; and the mean values of prothrombin timewere 9.61±0.74 s, 10.83±0.96 s, 10.14±1.03 s, 10.89±1.08 s and 11.07±1.03 s, respectively. At the statuses of hepatic normal, fibrosis and cirrhosis,the mean values of seralbumin were 30.11±3.46 g/L, 26.67±3.58 g/L and23.90±4.26 g/L, respectively; the mean values of serum bilirubin were 1.39±1.45 pmol/L, 6.87±3.96 pmol/L and 13.39±14.51 pmol/L, respectively;and the mean values of prothrombin time were 9.61±0.74 s, 10.60±1.04 sand 11.07±1.03 s, respectively. The hepatic function markers weresignificant difference among the piglets on different weekends or at differentstatuses of the disease during the course of modeling the disease (all p < 0.05).There were significant correlations of TTP with hepatic functional indexessuch as seralbumin, serum bilirubin and prothrombin time (r = -0.761, 0.666and 0.971, respectively) on different weekends, or with the statuses of chronichepatic diseases including hepatic normal, fibrosis and cirrhosis (r = -0.942,0.971 and 0.773, respectively). The diameter of SPV was correlated withseralbumin, serum bilirubin and prothrombin time on different weekends (r =-0.962, 0.682 and 0.802, respectively), or with the statuses of chronic hepaticdiseases including hepatic normal, fibrosis and cirrhosis (r = -1.0, 0.970 and0.950, respectively).Conclusion: (1) 40% carbon tetrachloride together with 5% alcohol(drinking water) could be used to successfully model hepatic fibrosis and livercirrhosis in piglets. (2) The TTP of portal vein on perfusion MRI and thediameters of SPV on MR portovenography could increase during the courseof modeling the disease. (3) The value of TTP of portal vein obtained byperfusion MRI and that of the SPV diameter are sensitive indicators fordetecting hepatic fibrosis and liver cirrhosis in piglets. (4) Standard magnetic resonance portal venography may serve as a non-invasive imaging modalityfor assessing the severity of hepatic fibrosis and liver cirrhosis. (5) The TTPof portal vein on perfusion MRI and the diameters of SPV on MRportovenography could be correlated with the laboratory hepatic functionalindexes, which suggested that MR portovenography might be used to evaluatethe hepatic function.