The Establishment Of Small Pig's Hepatic Fibrosis Model And Study With Multi-slice Computed Tomography Perfusion

Objective:To establish a model for small pig's hepatic fibrosis,by 16-sectioned spiral CT perfusion imaging was used to study dynamically the axiom for the liver's haemodynamic changes at different pathological periods, and calculus formula for the correlation of free portal venous pressure with priming volume and diagnosed hepatic fibrosis with MSCTP intraumatically.Materials and Methods:The subjects,supplied by the Animal Experiment Centre of Guangxi Medical University,were 45 Bama small pigs of the same germline,who were divided at random into a experiment group of 40 and a control group of 5,male or female,aged 4 months,weighing 10-13 kg.In the experiment,the establishment of the hepatic fibrosis model and the study with 16-slices spiral CT perfusion imaging were conducted simultaneously. Composite factors were employed to establish a model(a mixed feed of 65%of corn meal + 35%of Cholest + CC14 + PBS,with 10%absolute alcohol as the only one potion)for the experiment group,semel in die,16 weeks successively, and the normal feeding(65%of corn meal + 35%of rice bran,with clear water as the only one potion),was synchronically given to the controlled.The two groups of the animal received 16-slices spiral CT perfusion scanning, haemospasia for the biochemical indicators of hepatic function(incl.T.BIL, D.BIL,I.BIL,Alb,Glo,A/G,G-GT,ALT,AST,AST/ALT),laparotomy for measuring free portal venous pressure(FPP)and CT-aiming aspiration biopsy. The CT perfusion scanning,with Body-perfect multi-scanning procedure,firstly scanned with conventional axial view the liver,and then the deck of liver,spleen, aorta and portal vein for perfusion imaging scanning of the liver.Conditions for scanning:120kv,60mA,12mm/deck,0.5s/vic..And after auricular vein nyxis with 18^# transfixion pin and iohexol(300mg/ml),2ml/kg,flow rate of 2ml/s, given through high pressure syringe;and then with a lag of 4s,contrast agent was employed to have scanning for 60s successively.The perfusion data received image-analysis with functional software(Body perfect-syngo CT2007A) in the machine.CT(Hu)was determined respectively in the regions of interest of the aorta,portal vein,spleen and liver to acquire relative time density curve(TDC),i.e.correspondingly the density increase value from plain-scanning the regions of interest.A round or elliptic region of interest was taken in the aorta and portal vein,and the regions of interest in the liver and spleen were drawn along the organs' borderlines.The region of interest in the liver included mainly bulk of hepatic lobes at the scanned deck,with the liver's great vessels out of the way;the ROI of the spleen,the whole at the deck,was taken,some 1cm away from the organ's borderline'to avoid the volume effect.The software, with TDC produced automatically by it,calculated with the maximum slope rate method to produce automatically perfusion parameters:hepatic artery perfusion(HAP),portal venous perfusion(PVP),total hepatic blood flow (THBF)and hepatic artery perfusion index(HPI),of which THBF=HAP + PVR Each metered index was measured repeatedly twice,with the mean value taken. All the data obtained were analyzed statistically with SPSS13.0 software,the outcome of which were demonstrated through mean±standard deviation.The comparison of the multiple-grouped mean applied One-way ANOVA,and non-parameter rank sum test(Kruskal-Wallis test)for heterogeneity of variance, with P＜0.05 statistically different;double variable Person linear correlation analysis,respectively for four perfusion parameters.HAP,PVP,THBF and HPI of hepatic fibrosis grades 1-5,and the synchronic PVP measured value; regression equation for the derivation of the portal venous pressures at stages of hepatic fibrosis,with the results and the measured values given double var Pearson linear correlation analysis;detemined the case of pressure value induced with regression equation in 95%confidence interval,assessed the concidence of results tested by FPP and MSCTP respectively.Results:(1)No death in the control group of 5,with their liver's histopathologically normal appearing until the 16^thweek;14 deaths in the experiment group of 40 at the end of 4^thweek;2 deaths,the 8^thweek;models of 24 heads established.The death rate from model-building was 40%,with pathological models of stage 1,2,3,4 and 5 acquired respectively,of which 40 cases/vic,for Grade 0,27 for Grade 1,19 for Grade 2,17 for Grade 3,19 for Grade 4 and 16 for Grade 5.(2)Normal liver functional indexs:ALT(42.00±19.51)u/l,AST(27.60±9.71)u/l,AST/ALT (0.69±0.10),G-GT(55.40±33.14)u/l,Alb(31.46±3.96)g/l,Glo(19.74±9.12)g/l,A/G(1.12±0.05),T.BIL(4.30±1.19)umol/l,D.BIL(1.28±0.30)umol/l,I.BIL(3.024±0.89)umol/l.ALT,AST,AST/ALT,G-GT, Glo,D.BIL of the pathological models of 0-5 stages rose gradually,Alb,I.BIL decreased gradually and A/G ratio was inversion.All index in O grade in the normal control group and experimental group were not significant difference (p＞0.05).The interclass ALT,AST,G-GT,Alb,Glo,I.BIL were statistically significant(p＜0.05),while T.BIL,not statistically significant(p＞0.05).(3)FPP in normal control group was(5.00±0.61)mmHg.Pathologically staging 0-5 FPP stepped up gradually and were respectively:(5.01±0.77,6.85±1.10,10.68±1.19,14.09±4.16,17.02±3.67,18.33±3.75)mmHg,with interclass statistical significance(p＜0.05).All the indexs of O grade was not significant difference in normal control group and experimental group.(4)CT manifestation of the small pigs' hepatic fibrosis models:the normal small pigs and the experimented in grade O group,had slick liver surfaces,density uniformity,neither abnormal density focus in liver parenchyma seen,nor ascites formation,with proportionality of liver/spleen＞1.The livers' contours of the small pigs of grade 1 and 2,showed slight engorgement,acceptably slick surfaces,reduced liver's density,with CT proportionality of liver/spleen＜1,but no tumor formation in liver parenchyma,6 cases of which had evident manipulus ascites formation.The ones of grades 3 and 4 in the experiment group had insufficiency in slickness of livers' surfaces and in uniformed density of liver parenchyma,local densities reduced and evident focus-formed adiposis hepatica,but no solid tumor mass in the liver,with 11 cases of abdominal dropsy. The ones of grade 5 had slightly reduced liver's volume with unsmooth surface, seemingly evident nodosity shadow,uneven density in liver parenchyma without solid tumor mass formed,but 13 cases of ascites with no intumescent absorbent gland seen.(5)The HAP,PVP,HPI respectively was 27.74±11.77(ml.min^-1.100ml^-1);94.70±9.32(ml.min^-1.100ml^-1);THBF 22.44±8.79(ml.min^-1.100ml^-1);20.54±4.08(ml.min^-1.100ml^-1).CT perfusion parameters at pathological stages 0-5:HAP(28.24±11.74,18.02±4.70,24.24±9.62, 27.45±16.40,32.07±15.77,29.48±10.98)ml.min^-1.100ml^-1:stages 0-5PVP(91.69±15.44,77.73±10.52,61.82±10.52,45.08±14.98,36.18±6.22,32.39±6.15)ml.min^-1.100ml^-1:stages 0-5 THBF(119.93±21.87,95.76±12.09,86.06±11.65,72.54±27.00,68.26±18.91,62.54±13.65) ml.min^-1.100ml^-1;stages0-5HPI(22.53±6.25,28.48±5.58,30.82±6.05, 37.06±5.89,40.73±10.19,48.10±13.12)%.All perfusion indexs were not significant difference in O grade in normal control group and experimental group (P＞0.05).HAP of experimental first decreased and rose later,it was significant between 0 grade and 1 grade(P＜0.05).but not significant in 0 grade and 2,3,4,5 grade(P＞0.05);PVP and THBF in experimental group apt to decrease and had significant difference in all the group(P＜0.05)HPI in experimental group increased and significant in all the group(P＜0.05)(6)FPP in experimental group were in direct correlation with HAP,and HPI,(r=+0.263, +0.656)but inverse correlation with PVP and THBF,of which FPP was in the best correlation with PVP(r=-0.816).Multiple regression,Stepwise regression internalized F test statistics＜0.05 into regression equation,while F test statistics＞0.05 rejected from the equation.Of the 4 perfusion parameters stated above,HAP and THBF were rejected,and PVP and HPI,the two variables, internalized to establish the optima regression equation:regression equation for portal venous pressure and perfusion parameters Y=17.175-0.173X₁+0.135X₂ (Y:FPP X₁:PVP X₂:HPI).(7)The portal venous pressure was calculated with FPP and PVP,HPI linear regression equation,with the outcome as 12.7751±4.5631mmHg,the measured value 12.7867±5.4053mmHg, coefficient correlation:r=0.845,P＜0.01.We measured hepatic fibrosis at the 95%confidence interval.The results showed that the diagnosis rates of hepatic fibrosis were 84.78%,88.88%and 81.25%atâ… ,â…¡grade,â…¢,â…£grade, andâ…¤grade,respectively.Conclusion:1.The model-building from composite factor food of carbon tetrachloride(CCL4),agrypnal,absolute alcohol + high lipoids,low protein and low bilineurin can establish successfully the models of small pigs' liver fibrosis.2.Hepatic function test can cue damage to the liver,but unable to evaluate objectively pathological staging.3.With liver fibrosis,there is a correlation between the change of liver blood perfusion and the severity of the disease.4.The regression equation with PVP and HPI: Y=17.175-0.173X₁+0.135X₂(Y:FPP X₁:PVP X₂:HPI),was optimal for calculation of FPP in diagnosis hepatic fibrosis.