骨骼

ORIGINAL ARTICLE—LIVER,PANCREAS,AND BILIARY TRACTCombined measurements of serum bile acid level and splenic volume may be useful to noninvasively assess portal venous pressureHiromitsu Hayashi •Toru Beppu •Hirohisa Okabe •Hidetoshi Nitta •Katsunori Imai •Koichi Doi •Akira Chikamoto •Hideo BabaReceived:13February 2012/Accepted:2April 2012/Published online:9May 2012ÓSpringer 2012AbstractBackground We aimed to identify a noninvasive predic-tor of portal venous pressure (PVP).Methods We directly measured the PVP in 40consecu-tive patients who underwent direct percutaneous transhe-patic portal vein puncture as part of the therapeutic management for liver diseases,and we evaluated the association of the PVP with noninvasive clinical parame-ters.The backgrounds of the liver were normal in 13patients,chronic hepatitis in 17,and liver cirrhosis in ten.Results The mean PVP was 202±114mmH 2O.In a multivariate linear regression analysis,the serum bile acid level and splenic volume showed independent positive correlations with the PVP (P \0.001and 0.002,respec-tively).The formula for estimating PVP was identified as follows:PVP (mmH 2O)=serum bile acid (l mol/L)92.593?splenic volume (cm 3)90.416?65.929(R 2=0.698).In a receiver operating characteristic (ROC)anal-ysis,the AUC values of serum bile acid and splenic volume at a PVP of 200mmH 2O were 0.909and 0.758,respec-tively.However,the AUC values of serum bile acid and splenic volume at a PVP of 250mmH 2O were 0.792and 0.926,respectively,suggesting that the serum bile acid level and splenic volume are sensitive predictors of early and advanced portal hypertension,respectively.Conclusions Combined measurements of the serum bile acid level and splenic volume may be useful to noninva-sively assess the PVP prior to further invasive procedures.Keywords Serum bile acid ÁSplenic volume ÁPortal venous pressureAbbreviations ALT Alanine aminotransferase APRI Aspartate aminotransferase-to-platelet ratio index AST Aspartate aminotransferase AUC Area under the curve HVPG Hepatic venous pressure gradientICGR 15Indocyanine green retention rate at 15min NH 3Ammonia PeV Peripheral vein PVP Portal venous pressure ROC Receiver operating characteristic SMV Superior mesenteric vein SPV Splenic veinIntroductionMany conditions,of which liver cirrhosis is the most common,can result in portal hypertension.Portal hyper-tension leads to unfavorable findings such as the develop-ment of ascites,splenomegaly with pancytopenia,and portosystemic collaterals.Portosystemic collaterals directly connect the portal blood vessels to the general circulation–bypassing the liver–and this results in portosystemic encephalopathy and the formation of esophageal and gas-tric varices.The formation of esophageal and gastric var-ices is usually without any symptoms until the varices rupture,and massive bleeding from the varices may cause life-threatening bleeding.For treating portal hypertension,non-selective beta-blockers have been used to reduce the portal venous pressure (PVP),and the reduction in PVPH.Hayashi ÁT.Beppu ÁH.Okabe ÁH.Nitta ÁK.Imai ÁK.Doi ÁA.Chikamoto ÁH.Baba (&)Department of Gastroenterological Surgery,Graduate School of Medical Sciences,Kumamoto University,1-1-1Honjo,Kumamoto 860-8556,Japane-mail:hdobaba@kumamoto-u.ac.jpJ Gastroenterol (2012)47:1336–1341DOI 10.1007/s00535-012-0599-7was found to be higher in patients without varices[1]. Thus,the early detection of portal hypertension would enable us to diagnose esophageal and gastric varices before they rupture and to assess treatment before the formation of a severe portosystemic shunt[2,3],indicating that the assessment of the portal hypertension in the initial stage is clinically meaningful.However,the direct measurement of PVP,done by puncturing the portal vein either through a percutaneous-transhepatic or transjugular approach,is an invasive procedure.Nowadays,although hepatic venous pressure gradient(HVPG)measurement has been found to be useful for assessing the severity of portal hypertension, the routine and frequent use of HVPG for PVP assessment is not recommended because of its relatively high inva-siveness,cost,and the need for technical expertise[4]. Here,we evaluate the association between noninvasive clinical parameters and direct PVP,and from ourfindings we suggest that combined measurements of the serum bile acid level and splenic volume may be useful to noninva-sively assess the PVP.Patients,materials,and methodsSubjectsFrom May2006to November2008,40consecutive patients underwent direct percutaneous transhepatic portal vein puncture as part of the therapeutic management of their liver diseases;the procedures were performed at the Department of Gastroenterological Surgery,Graduate School of Medical Sciences,Kumamoto University.Direct portal vein puncture was performed during portal vein embolization as preoperative management for38patients. Transportal chemoembolization was performed in1patient for treating a rare type of hepatocellular carcinoma that was mainly supplied by portal bloodflow.Shunt emboli-zation was performed in1patient for a massive arterial–portal shunt;this patient had the highest PVP in the cohort. We routinely measured the PVP by using a water manometer after direct portal vein puncture,and the PVP value was used for analysis.The subjects consisted of34 men and6women.All40patients had liver tumors,and none had idiopathic portal hypertension.Portal vein puncture was indicated for patients with hepatocellular carcinoma(n=30),intrahepatic cholangiocarcinoma (n=3),metastatic liver tumor of colon cancer(n=4), hemangioma(n=1),and bile duct carcinoma(n=2). None of the patients had obstructive jaundice.Hepatitis B surface antigen was detected in four patients,hepatitis C antibody in18,and neither of them(nonB,nonC)in18 patients.The backgrounds of the liver were proven path-ologically in30patients;in the others the diagnosis was established by a combination of laboratory data and radiologicalfindings.The backgrounds of the liver were normal in13patients,chronic hepatitis in17,and liver cirrhosis in ten.The other clinical characteristics of these 40patients are presented in Table1.Written informed consent was obtained from all patients,and this trial was approved by the ethics committee of our institute.Measurement of portal venous pressureThe safest and most suitable portal vein to be punctured was selected by ultrasonographic exploration.After local anes-thesia was administered,the portal vein was punctured using an18-gauge needle under continuous ultrasonographic guid-ance.Next,a7-Fr vascular sheath was inserted into the portal vein.Underfluoroscopic guidance,a7-Fr catheter(Selecon MP catheterÒ;Clinical Supply,Gifu,Japan)was advanced into the main portal vein,and the position for the measurement was confirmed by the use of a small amount of contrast medium,as previously reported[5,6].The catheter tip was flushed with saline,and the PVP was then obtained by con-necting the catheter to a water manometer.In all measure-ments,the external zero reference point was set at the midaxillary line of the patient,as previously described[7,8]. Portal hypertension is defined as a state where portal vein pressure is more than200mmH2O[9].The procedure was well tolerated in all patients.Clinical parameters and splenic volumetryThe following noninvasively available clinical data were also obtained from each patient before portal vein Table1Clinical characteristics in40patientsVariables Mean±SD RangeAge(years)68.8±9.344–84 Platelet count(9104/l L)16.2±7.9 5.2–41.3 Serum albumin(g/dL) 3.83±0.41 2.6–4.7 Total bilirubin(mg/dL)0.81±0.330.3–1.8 AST(IU/L)47.1±29.316–149 ALT(IU/L)46.7±33.39–120 Prothrombin activity(%)89.8±16.630–124 NH3(l g/dL)45.7±25.010–119 Hyaluronic acid(ng/mL)165±13822–664 Serum bile acid(l mol/L)21.6±22.8 1.9–95 ICGR15(%)17.3±10.6 2.0–46.4 Splenic volume(cm3)190±10952–612 APRI 1.21±1.300.11–6.56AST aspartate aminotransferase,ALT alanine aminotransferase,NH3 ammonia,ICGR15indocyanine green retention rate at15min,APRI aspartate aminotransferase-to-platelet ratio indexpuncture:age,platelet count;prothrombin activity;indo-cyanine green retention rate at15min(ICGR15);and blood levels of total bilirubin,aspartate aminotransferase(AST), alanine aminotransferase(ALT),serum albumin,ammonia (NH3),and serum hyaluronic acid.The splenic volume was measured by computed tomographic volumetry with con-trast enhancement,as described previously[10,11].The aspartate aminotransferase-to-platelet ratio index(APRI) was calculated using the following formula,as described previously[12]:APRI=AST level(/upper limit of normal)/platelet countsð109=LÞÂ100The upper limit of the normal AST value at our institute was34IU/L.Collection of serum samples in the peripheral,superior mesenteric,and splenic veins and measurementof the total bile acid level in serumPeripheral serum samples were obtained before portal vein puncture in all40patients,who had fasted overnight before the procedure.In37of the40patients,we compared the serum bile acid levels in the peripheral vein(PeV),superior mesenteric vein(SMV),and splenic vein(SPV).After the measurement of the PVP,the catheter tip was advanced into the SPV(near the hilum of the spleen)and the SMV underfluoroscopic guidance.The position of the catheter tip was confirmed using a small amount of contrast med-ium.In each vein,whole blood was obtained and drawn using a sterile syringe,transferred to a centrifuge tube, allowed to clot,and centrifuged(3000rpm for15min). The separated serum was frozen at-80°C until the assay. The serum total bile acid level was determined at the Department of Diagnostic Medicine,Kumamoto Univer-sity,using an Aqua Auto Kainos TBA reagent(Kainos, Tokyo,Japan)according to the manufacturer’s instructions. Statistical analysesStatistical analyses were performed using a commercial statistical software package(SPSS for Windows,version 11.0;SPSS,Chicago,IL,USA).Continuous values were evaluated using the Mann–Whitney U-test.To analyze correlations between PVP and clinical variables and to determine a formula for estimating PVP,univariate and stepwise multivariate linear regression analyses were used.A receiver operating characteristic(ROC)curve was used to determine suitable cutoff values of clinical variables that yielded the highest combined sensitivity and specificity.A P value of less than0.05(two-tailed t-test)was considered statistically significant.ResultsPortal venous pressureThe mean PVP in the40patients was202±114mmH2O. The PVP was elevated to more than200mmH2O in14 patients.The PVP was equal to or less than200mmH2O in 26patients.A histogram of the PVP values in the40 patients is shown in Fig.1.Correlation between the portal venous pressureand noninvasive clinical factorsCorrelations between13noninvasive clinical factors and the PVP were analyzed using a linear regression analysis (Table2).In the univariate analysis,the platelet count was shown to have a significant negative correlation with the PVP.In contrast,AST,NH3,hyaluronic acid,APRI,serum bile acid,ICGR15,and splenic volume showed a significant positive correlation with the PVP.In the multivariate analysis,the serum bile acid and splenic volume showed independent positive correlations with the PVP.The for-mula for estimating the PVP was determined from the linear regression analysis and is as follows:Portal venous pressureðmmH2OÞ¼serum bile acidðl mol=LÞÂ2:593þsplenic volumeðcm3ÞÂ0:416þ65:929R2¼0:698ÀÁROC analyses of serum bile acid level and splenic volume at portal venous pressures of200and250mmH2OTo determine a suitable cutoff value for the serum bile acid level and splenic volume that yielded the highest combined sensitivity and specificity at PVPs of200and250mmH2O, ROC analyses were performed(Table3).The values for the area under the curve(AUC)of the serum bile acid andsplenic volume at a PVP of200mmH2O were0.909and 0.758,respectively.However,the AUC values of the serum bile acid and splenic volume at a PVP of250mmH2O were 0.792and0.926,respectively.Suitable cutoff values of the serum bile acid level and splenic volume that yielded the highest combined sensitivity and specificity with respect to discriminating between patients with a PVP of more than 200mmH2O and those with a PVP of200mmH2O or less were20and200mL,respectively.The suitable cutoff values of the serum bile acid level and splenic volume at a PVP of250mmH2O were22.75l mol/L and250mL, parison of serum total bile acid levelsin the peripheral,superior mesenteric,and splenic veins To further assess the implications of the positive correla-tion between peripheral serum bile acid level and PVP,we investigated the serum bile acid levels in the PeV and portal system(SMV and SPV).In37patients,we com-pared the serum bile acid levels in the PeV,SMV,and SPV (Fig.2).The median serum bile acid levels(ranges)in the PeV,SMV,and SPV in these37patients were13.3 (1.9–95)l mol/L,25.3(1.3–502.7)l mol/L,and 6.6 (0.4–181.1)l mol/L,respectively.The serum bile acid level was highest in the SMV and lowest in the SPV.The serum bile acid levels in the SMV were significantly higher than that in the PeV and SPV(P=0.003and\0.001, respectively).In addition,the serum bile acid level in the PeV was significantly higher than that in the SPV (P=0.001).Thus,there were significant difference in the serum bile acid level among the PeV,SMV,and SPV (SMV[PeV[SPV).Next,we examined differences in the serum bile acid levels in the PeV,SMV,and SPV in patients with a PVP of B200mmH2O(without portal hypertension)and those with a PVP of[200mmH2O(portal hypertension) (Fig.3),because portosystemic shunt,which is induced by portal hypertension,has been reported to be associated with an elevation in serum bile acids[13,14].In the23patients with PVP B200mmH2O,the median levels(ranges)ofTable2Simple and multiple stepwise linear regression analyses of clinical variables and portal venous pressureVariables Univariateanalysis Multivariate analysis(stepwise)b P value b P valueAge(years)-0.2490.121–Platelet count(9104/l L)-0.4600.003–Serum albumin(g/dL)-0.2920.068–Total bilirubin(mg/dL)0.0810.621–AST(IU/L)0.4120.008–ALT(IU/L)0.4440.004–Prothrombin activity(%)-0.4500.004–NH3(l g/dL)0.5060.001–Hyaluronic acid(ng/mL)0.628\0.001–Serum bile acid(l mol/L)0.781\0.0010.516\0.001 ICGR15(%)0.3500.029–Splenic volume(cm3)0.743\0.0010.3990.002 APRI0.633\0.001–b standardized regression coefficient,AST aspartate aminotransferase, ALT alanine aminotransferase,NH3ammonia,ICGR15indocyanine green retention rate at15min,APRI aspartate aminotransferase-to-platelet ratio indexPortal venous pressure(mmH2O)=serum bile acid(l mol/L)9 2.593?splenic volume(cm3)90.416?65.929(R2=0.698) Table3ROC analyses of serum bile acid levels and splenic volumes at portal venous pressures of200and250mmH2O(n=40) Variables200mmH2O250mmH2OSerum total bile acid(l mol/L)2022.75 AUC0.9090.792 Sensitivity(%)85.785.7 Specificity(%)88.575.8 Splenic volume(cm3)200250AUC0.7580.926 Sensitivity(%)71.485.7 Specificity(%)84.693.9ROC receiver operating characteristic,AUC area under the curveserum bile acid in the PeV,SMV,and SPV were7.6 (1.9–31.2)l mol/L,12.2(1.3–216.6)l mol/L,and 3.8 (0.4–26.3)l mol/L,respectively.In these patients,the serum total bile acid level in the SMV was significantly higher than the levels in the PeV and SPV(P=0.004and \0.001,respectively).In patients with PVP B200mmH2O, the bile acid level in the PeV was significantly higher than that in the SPV(P=0.013).In the14patients with PVP [200mmH2O,the median levels(ranges)of serum bile acid in the PeV,SMV,and SPV were36.2(5.0–95)l mol/L,44.6 (13.6–502.7)l mol/L,and18.8(2.6–181.1)l mol/L,respec-tively.In patients with PVP[200mmH2O,although the serum bile acid level in the SMV was markedly higher than that in the SPV(P=0.001),no significant difference in serum bile acid levels was found between the PeV and the SMV(P=0.167).Thus,in patients without portal hyper-tension(PVP B200mmH2O),the serum bile acid level in the SMV was markedly higher than that in the PeV,whereas in the patients with portal hypertension,the serum bile acid level in the PeV was close to that in the SMV.DiscussionIn this study,the serum bile acid level and the splenic volume showed independent positive correlations with the PVP.It was considered that the monitoring of the serum bile acid level and splenic volume could be useful to noninvasively assess the occurrence and progress of portal hypertension.The formula for estimating portal venous pressure(PVP)was identified as follows:PVP (mmH2O)=serum bile acid(l mol/L)92.593?splenic volume(cm3)90.416?65.929(R2=0.698).Further-more,the AUC of serum bile acid at a PVP of200mmH2O was higher than that of splenic volume(0.909and0.758, respectively),whereas the AUC of serum bile acid at a PVP of250mmH2O was lower than that of splenic volume (0.792and0.926,respectively).These results suggest that the serum bile acid level and splenic volume are sensitive predictors of early and more advanced portal hypertension, respectively.The combined measurements of the serum bile acid level and splenic volume could be clinically and economically useful to distinguish those patients who truly require further assessment for portal hypertension by using more invasive and expensive procedures such as HVPG.In the present study,esophageal varices were found in seven patients,and the lowest PVP was215mmH2O(data not shown).In patients with a serum bile acid level of more than20l mol/L,periodical endoscopic variceal screening may be necessary.The SMV is known to be the main route of transport for intestinally absorbed bile acids.In several previous studies [13,14],a portosystemic shunt has been reported to be associated with increased peripheral serum bile acid levels. For example,an elevated serum bile acid level in the PeV was positively associated with the degree of portal and splenic vein shunts in cirrhotic patients[13].In addition,a portosystemic shunt has been reported to induce elevations in serum bile acid levels without any changes in such liver function parameters as serum albumin and bilirubin levels and prothrombin activity[14].Experimentally,in normal rats and dogs,an end-to-side portocaval shunt has also been reported to increase the peripheral bile acid level in serum [15,16].To further assess the implications of the increased peripheral serum bile acid level in portal hypertension,we investigated the serum bile acid levels in the PeV and the portal system(SMV and SPV).Interestingly,there were significant differences in the serum bile acid levels amongthe PeV,SMV,and SPV(SMV[PeV[SPV).In patients without portal hypertension(PVP B200mmH2O),the serum bile acid level in the SMV was markedly higher than that in the PeV,whereas in patients with PVP [200mmH2O,the serum bile acid level in the PeV was close to that in the SMV(i.e.,the difference in serum bile acid levels between the PeV and SMV was not significant). On the basis of our results and previousfindings,we speculate that the leaking of SMV-derived blood contain-ing high levels of serum bile acid through a portosystemic shunt may be one of the causes of the elevated peripheral serum bile acid levels in patients with portal hypertension.In the present study,the splenic volume,as well as the serum bile acid level,was independently and positively associated with the PVP.Of note,a splenic volume greater than200cm3 was associated with a PVP greater than200mmH2O.Indeed, the mean normal splenic volume,determined by computed tomographic volumetry in150healthy adults,was reported to be112cm3(range32–209cm3)[17].Although the criterion for splenomegaly in regard to splenic volume is not yet defined, a splenic volume of200cm3may be a useful criterion for splenomegaly caused by portal hypertension.A limitation of the present study was that all our patients had various types of liver tumors,ranging from malignant to benign.In patients with normal portal pressure and without liver tumors,it would be rare to perform direct portal puncture.If possible,the usefulness of our formula should be re-evaluated and modified in further studies dealing with patients without liver tumors.In conclusion,peripheral serum bile acid level and splenic volume were independently and positively correlated with PVP,suggesting the usefulness of these parameters as non-invasive predictors of PVP.The monitoring of the serum bile acid level and splenic volume may be clinically useful for the early detection and management of portal hypertension to distinguish those patients who require further invasive and expensive procedures such as HPVG.Conflict of interest The authors declare that they have no conflict of interest.References1.Escorsell A,Ferayorni L,Bosch J,Garcia-Pagan JC,Garcia-TsaoG,Grace ND,Rodes J,et al.The portal pressure response to beta-blockade is greater in cirrhotic patients without varices than in those with varices.Gastroenterology.1997;112:2012–6.2.Sharma SK,Aggarwal R.Prediction of large esophageal varicesin patients with cirrhosis of the liver using clinical,laboratory and imaging parameters.J Gastroenterol Hepatol.2007;22:1909–15.3.Kadouchi K,Higuchi K,Shiba M,Okazaki H,Yamamori K,Sasaki E,Tominaga K,et al.What are the risk factors for aggravation of esophageal varices in patients with hepatocellular carcinoma?J Gastroenterol Hepatol.2007;22:240–6.4.Thabut D,Moreau R,Lebrec D.Noninvasive assessment ofportal hypertension in patients with cirrhosis.Hepatology.2011;53:683–94.5.Beppu T,Iwatsuki M,Okabe H,Okabe K,Masuda T,Hayashi H,Sugiyama S,et al.A new approach to percutaneous transhepatic portal embolization using ethanolamine oleate iopamidol.J Gas-troenterol.2010;45:211–7.6.Beppu T,Hayashi H,Okabe H,Masuda T,Mima K,Otao R,Chikamoto A,et al.Liver functional volumetry for portal vein embolization using a newly developed99mTc-galactosyl human serum albumin scintigraphy SPECT-computed tomography fusion system.J Gastroenterol.2011;46:938–43.7.Kallio H.Portal pressure and collaterals in cirrhotic patients.AnnChir Gynaecol.1984;73:308–12.8.Christensen U,Sorensen TI,Jensen LI,Aagaard J,Burcharth F.The free portal pressure in awake patients with and without cir-rhosis of the liver.Liver.1983;3:147–50.9.Segawa M,Sakaida I.Diagnosis and treatment of portal hyper-tension.Hepatol Res.2009;39:1039–43.10.Hayashi H,Beppu T,Okabe K,Masuda T,Okabe H,Baba H.Risk factors for complications after partial splenic embolization for liver cirrhosis.Br J Surg.2008;95:744–50.11.Hayashi H,Beppu T,Okabe K,Masuda T,Okabe H,Ishiko T,Baba H.Therapeutic factors considered according to the preop-erative splenic volume for a prolonged increase in platelet count after partial splenic embolization for liver cirrhosis.J Gastroen-terol.2010;45:554–9.12.Wai CT,Greenson JK,Fontana RJ,Kalbfleisch JD,Marrero JA,Conjeevaram HS,Lok AS.A simple noninvasive index can predict both significantfibrosis and cirrhosis in patients with chronic hepatitis C.Hepatology.2003;38:518–26.13.Ohkubo H,Okuda K,Iida S,Ohnishi K,Ikawa S,Makino I.Roleof portal and splenic vein shunts and impaired hepatic extraction in the elevated serum bile acids in liver cirrhosis.Gastroenter-ology.1984;86:514–20.14.Poupon RE,Poupon RY,Grosdemouge ML,Erlinger S.Effect ofportacaval shunt on serum bile acid concentration in patients with cirrhosis.Digestion.1977;16:138–45.15.Herz R,Paumgartner G,Preisig R.Bile salt metabolism and bileformation in the rat with a portacaval shunt.Eur J Clin Invest.1974;4:223–8.16.Horak W,Gangl A,Funovics J,Grabner G.Effect of portacavalshunt and arterialization of the liver on bile acid metabolism.Gastroenterology.1975;69:338–41.17.Kaneko J,Sugawara Y,Matsui Y,Ohkubo T,Makuuchi M.Normal splenic volume in adults by computed tomography.Hepatogastroenterology.2002;49:1726–7.。