Wedged Hepatic Venous Pressure Adequately Reflects Portal Pressure in Hepatitis C Virus Related Cirrhosis

Transcription

1 Wedged Hepatic Venous Pressure Adequately Reflects Portal Pressure in Hepatitis C Virus Related Cirrhosis ANTONIA PERELLÓ, 1 ÀNGELS ESCORSELL, 1 CONCEPCIÓ BRU, 2 ROSA GILABERT, 2 EDUARDO MOITINHO, 1 JUAN CARLOS GARCíA-PAGÁN, 1 AND JAUME BOSCH 1 Wedged hepatic venous pressure (WHVP) is equivalent to portal venous pressure in patients with alcoholic liver diseases. However, it may underestimate portal pressure in nonalcoholics, which is important because hepatitis C virus (HCV) infection is a frequent cause of chronic liver disease. We investigated the agreement between directly measured portal pressure and WHVP in alcoholic and HCV related liver diseases. Seventy-one patients with liver disease resulting from HCV infection (n 32), alcohol (n 25), or both (n 14) underwent simultaneous measurements of WHVP (by hepatic vein catheterization) and portal pressure (by direct puncture). In 9 patients, measurements were repeated 20 minutes after acute iv propranolol administration. WHVP showed an excellent agreement with portal pressure in patients with cirrhosis resulting from either HCV, alcohol or both (intraclass correlation coefficient: 0.94, 0.93, and 0.97, respectively; P F.001). A discrepancy of H5 mm Hg was observed in 7 cases. WHVP underestimated portal pressure in only 1 case and exceeded portal pressure by H5 mm Hg in 6 patients. The WHVP response to propranolol closely and significantly correlated with changes in portal pressure (intraclass correlation coefficient: 0.87; P F.004). The simple and safe measurement of WHVP accurately reflects portal pressure in alcoholic and HCV related liver disease. This technique also allows us to accurately assess the portal pressure response to propranolol in both alcoholic and HCV related cirrhosis. (HEPATOL- OGY 1999;30: ) Measurements of portal venous pressure (PP) are increasingly used in the evaluation of patients with chronic liver diseases because they provide useful prognostic information regarding the risk of variceal bleeding, 1-4 the outcome of Abbreviations: PP, portal venous pressure; HVPG, hepatic venous pressure gradient; WHVP, wedged hepatic venous pressure; FHVP, free hepaticvenous pressure; PBC, primary biliary cirrhosis; HCV, hepatitis C virus; TIPS, transjugular intrahepatic portosystemic shunt; PPG, portal pressure gradient; Ri, intraclass correlation coefficient; r, Spearman correlation coefficient; HBV, hepatitis B virus; MAP, mean arterial pressure. From the 1 Hepatic Hemodynamic Laboratory, Liver Unit, Institut de Malalties Digestives, and 2 Centre de Diagnòstic per la Imatge, Hospital Clínic, IDIBAPS, University of Barcelona, Barcelona, Spain. Received June 28, 1999; accepted September 28, Antonia Perelló was recipient of a grant from the Programa Nacional de Formaciónde Personal Investigador, Ministerio de Educación y Ciencia (PN95/ ). Address reprint requests to: Jaume Bosch, Hepatic Hemodynamic Laboratory, Liver Unit, Hospital Clínic, C/ Villarroel, 170, Barcelona, Spain. jbosch@ medicina.ub.es; fax: Copyright 1999 by the American Association for the Study of Liver Diseases /99/ $3.00/ patients with acute variceal hemorrhage, 5,6 and long-term survival The direct measurement of PP is an invasive technique that requires the direct puncture of the portal vein, either through a percutaneous or transjugular approach. Furthermore, an additional procedure is frequently required to measure the inferior vena cava pressure, which is necessary to calculate the porto-caval pressure gradient. Because of these limitations, the safe and reproducible technique of hepatic vein catheterization, with measurements of the hepatic venous pressure gradient (HVPG) the difference between wedged (WHVP) and free (FHVP) hepatic venous pressures has become the preferred technique for the evaluation of portal hypertension. Many studies have shown that in alcoholic liver disease, WHVP equals portal pressure (sinusoidal portal hypertension) Because FHVP is equivalent to inferior vena cava pressure, the HVPG equals the porto-caval pressure gradient in patients with sinusoidal portal hypertension. 15 Previous studies in nonalcoholic liver cirrhosis 11,16 or chronic active hepatitis 12 suggested that, in a substantial proportion of patients, WHVP may underestimate PP. Although it has been shown that WHVP adequately reflects PP in HBV related cirrhosis, 17 it is well documented that patients with primary biliary cirrhosis (PBC) 18 frequently exhibit a presinusoidal component in their portal hypertension. Finally, there is no data with regards to cirrhosis related to HCV infection, which is becoming the main cause of chronic liver disease and portal hypertension in Western countries. 19 Moreover, alcoholics frequently have coincidental HCV serum antibodies 20,21 which, in the absence of other markers, makes it difficult to decide which is the primary cause of the chronic liver disease. Measurements of HVPG are increasingly used to assess the effects of continued pharmacological therapy for portal hypertension. 1-3,14 However, studies comparing changes in PP with those observed in WHVP during propranolol administration are scarce and yield conflicting results, 13,14,22 adding uncertainty as to the use of hepatic vein catheterization to assess portal hypertension in patients with HCV related liver diseases. The present study was designed to address these issues by performing simultaneous measurements of WHVP and PP in a large series of patients with chronic liver disease, either with or without HCV infection. MATERIALS AND METHODS The study was performed in 76 alcoholic and/or HCV infected patients, referred to the Hepatic Hemodynamic Laboratory for

2 1394 PERELLÓ ET AL. HEPATOLOGY December 1999 evaluation or treatment of portal hypertension. The presence of liver disease was assessed by clinical, biochemical, and ultrasonographic parameters. Liver biopsy was obtained whenever possible, including all cases without clinical signs of decompensated cirrhosis and portal hypertension (ascites, variceal hemorrhage, hepatic encephalopathy). The presence of other etiologies was excluded by HBV testing and measurements of -1-antitripsin, serum and urinary copper, ceruloplasmin, serum iron, iron-binding capacity, transferrin, immunoglobulins, and serum autoantibodies. Patients were etiologically classified in 3 groups according to the history of ethanol consumption and serological tests for HCV. Alcoholism was defined as a continuous intake of alcoholic beverages with an average ethanol consumption of greater than 80 g per day. All patients had an upper gastrointestinal endoscopy to assess the presence of gastroesophageal varices. The severity of liver dysfunction was evaluated according to the Child-Pugh score. No patient was receiving vasoactive drugs at the time of the study. All patients gave written, informed consent to participate in the study, which was approved by the Clinical Research Committee of Hospital Clínic in May Thirty-four patients had chronic liver disease resulting from HCV infection, 28 had alcoholic liver disease, and 14 were alcoholics with an HCV infection. Five initially enrolled patients (2 HCV infected and 3 alcoholics) were excluded because of complete portal vein thrombosis (n 1), presence of veno-venous communications precluding adequate occlusion of the hepatic vein for measurement of WHVP (n 1) and vagal reactions when attempting the portal vein puncture (n 3). Thus, 71 patients were finally studied: 32 patients with HCV related liver disease, 25 alcoholics, and 14 patients with both conditions. Table 1 shows the clinical and hemodynamic data of the patients included in the study: 56 patients were men and 15 women, with a mean age of years (mean standard deviation (SD); range: 28 to 80 years). Gastroesophageal varices were present in 57 patients and ascites in 39. Nine patients participating in the study (6 male and 3 female, with a mean age of years) had repeated measurements of WHVP and PP before, and 20 minutes after, the acute -adrenergic blockade by the iv administration of propranolol (0.15 mg/kg). The cause of liver disease was HCV infection in 3 cases, alcoholism in 3, and HCV plus alcohol in the remaining 3 cases. Six of these 9 patients had esophageal varices, and none had ascites. Before the hemodynamic measurements, all patients underwent duplex-doppler ultrasonography. Special attention was paid to the presence of signs of portal hypertension, to the patency of the portal vein, to the direction of portal blood flow, and to the presence of grossly dilated collaterals. Measurement of WHVP, FHVP, and HVPG. These parameters were measured during hepatic vein catheterization following previously TABLE 1. Characteristics of the Patients Included in the Study According to the Cause of Liver Disease HCV infection n 32 Alcohol n 25 HCV Alcohol n 14 Sex (male/female) 20/12 23/2 13/1 Age (years)* Child-Pugh score* Presence of varices (n) Presence of ascites (n) MAP (mmhg)* Heart rate (bpm)* Portal pressure (mmhg)* Portal pressure gradient (mmhg)* WHVP (mmhg)* FHVP (mmhg)* HVPG (mmhg)* *Data expressed as mean SD. P.05 vs. the 2 other groups. described methods. 15 After an overnight fast, the patients were transferred to the Hepatic Hemodynamic Laboratory where, under local anesthesia, a catheter introducer (USCI International, Galway, Ireland) was placed into the right femoral or the right internal jugular vein using the Seldinger technique. It was used to introduce a 7F balloon-tipped catheter (Meditech, Cooper Scientific Corp, Watertown, MA), which was guided to the main right hepatic vein under fluoroscopic control. WHVP is measured by occluding the hepatic vein, which can be done by either advancing the catheter as far as it can go to have it wedged in a hepatic vein radicle or, preferably, by inflating the balloon at the tip of the catheter, thus occluding the hepatic vein at approximately 3 to 5 cm from its opening into the inferior vena cava. 23 At this moment, the stagnant column of blood transmits the pressure at the hepatic sinusoids (hepatic sinusoidal pressure). The balloon occlusion technique was used in all the patients included in this study. The adequate occlusion of the hepatic vein was confirmed by lack of contrast reflux following the hand injection of a small amount of contrast medium. 15,23 Once the catheter does not occlude the hepatic vein, the free pressure (FHVP) is obtained. Measurement of Portal Venous Pressure. Portal venous pressure was measured by either percutaneous puncture under ultrasonographic and radiological guidance (n 25) or by a transjugular approach during the placement of a transjugular intrahepatic portosystemic shunt (TIPS) (n 46). The portal pressure gradient (PPG) was calculated as the portal venous pressure, minus the FHVP, measured in the same conditions described for the measurement of WHVP. Percutaneous Approach. While the hepatic vein catheter was maintained in the hepatic vein to measure the hepatic venous pressures, the portal vein was punctured percutaneously at the same time, using a thin needle (20G) under local anesthesia and continuous ultrasonographic guidance. The safest and most suitable branch of the portal tract was selected by ultrasonographic exploration, choosing the nearest branch to the abdominal wall. The correct position for the measurement was confirmed by the sonographic image and by a gentle aspiration of portal venous blood. The needle was then flushed with 3 ml of saline and connected to another pressure transducer to measure portal venous pressure. Atropine (0.5 mg, intramuscular) and meperidine (50 mg, intramuscular) were administered before portal vein puncture to avoid vagal reactions (which nevertheless occurred in 3 patients, who were discontinued from the study and recovered uneventfully). There were no other complications, such as bile leak or intra-abdominal bleeding. Transjugular Approach. This was used in patients undergoing a TIPS procedure. Twelve patients underwent emergency TIPS because of uncontrolled variceal bleeding; 29 had elective TIPS after repeated variceal bleeding, despite pharmacological and endoscopic therapy; and the remaining 5 cases had TIPS because of refractory ascites. The right portal vein was catheterized and punctured, as previously described 24 with a transjugular liver biopsy needle advanced through a long sheath venous introducer and guiding catheter (Cook set TJL-100-U-EC-18, EC-19; Bjaeverskov, Denmark) under continuous ultrasonographic and radiological guidance. Patients were consciously sedated with midazolam (0.1 mg/kg, iv) and meperidine ( mg, iv). Immediately after the portal vein puncture, the needle was retired, and a 5F catheter was advanced into the main portal trunk where the portal venous pressure was simultaneously measured with the FHVP. The latter was measured through the side arm of the introducer sheath, which was advanced 2 cm into the hepatic vein. In all measurements, the external zero reference point was set at the midaxilary line of the patient. All venous pressures (WHVP, FHVP, and PP) were measured at least by duplicate, for a period of 40 or more seconds, by means of precalibrated, high sensitivity pressure transducers, connected to a multichannel recorder (Hewlett- Packard A, Hewlett-Packard Co., Waltham, MA). Permanent tracings were obtained and read by an experienced investigator who

3 HEPATOLOGY Vol. 30, No. 6, 1999 PERELLÓ ET AL did not know any data about the patient. The results reported represent the mean of the duplicate or triplicate measurements, averaged to the closer half mm Hg (thus, a triplicate reading of 20, 20.5, and 20.5 mm Hg would be reported as 20.5 mm Hg, not as mm Hg). Heart rate, systolic, diastolic, and mean arterial pressure were monitored throughout the study by using an automatic vital signs monitor (Dinamap, Critikon Inc, Tampa, FL). Effects of Acute Propranolol Administration. As previously mentioned, a subgroup of 9 patients undergoing simultaneous measurements of WHVP and PP by the percutaneous approach, had a second set of pressure measurements 20 minutes after the acute iv administration of 0.15 mg/kg of propranolol. Statistical Analysis. All values are reported as mean SD. Comparisons within each group were done by using the two-tailed Student s t test. Comparisons between groups were done by using the analysis of variance with the Bonferroni correction when required, and using the 2 -test, with the Yates correction, to analyze qualitative data. Agreement between measurements was assessed by using both the Spearman s correlation coefficient (r) and the intraclass correlation coefficient (Ri). 25 When using the latter, the agreement was considered to be excellent when Ri was.75, good when Ri was between.75 and.60, and nonacceptable when Ri was.60. Ri was calculated as follows: Ri (SD ti ) 2 (SD di ) 2 / (SD ti ) 2 (SD di ) 2 2/n [n.di 2 (SD di ) 2 ] where ti was the mean value of PP WHVP of each patient; SD ti was the standard deviation of ti; di was the mean value of PP WHVP for each individual patient; SD di was the standard deviation of di; and n was the number of paired measurements. We considered that there was a discrepancy between PP and WHVP when these measurements differed by 5 mm Hg or more for any individual patient. The statistical SPSS package (SPSS Inc., Chicago, IL) was used for the analysis. Statistical significance was established at a P value of less than.05. RESULTS Seventy-one patients were included in the study. Among them, 32 patients had HCV related liver disease: 26 patients had cirrhosis, 5 had chronic hepatitis, and 1 had only minimal changes at liver biopsy. The alcoholic liver disease group included 23 cirrhotic patients and 2 patients with steatosis at liver biopsy. All of the 14 patients with both HCV infection and chronic alcohol abuse had cirrhosis. As shown in Table 1, HCV infected patients were older than patients belonging to the other 2 groups. The alcoholic patients had a more advanced liver disease, as shown by a higher Child- Pugh score and clinical manifestations of portal hypertension (Table 1). All but 3 patients (2 alcoholics with steatosis and 1 HCV positive patient with minimal changes on liver biopsy) had portal hypertension, defined by a portal pressure gradient (PPG) greater than 5 mm Hg. The PPG averaged mm Hg in patients with HCV infection; mm Hg in alcoholic patients; and mm Hg in patients with both alcoholism and HCV infection (ns) (Table 1). Agreement Between Portal Venous Pressure and WHVP. As shown in Fig. 1, the concordance between PP and WHVP for the global series was excellent (r.95, Ri.99; P.00001). The same holds true when the 3 subgroups of patients with HCV infection, chronic alcoholism, or both were analyzed separately (r values:.95,.91, and.92, respectively; Ri values:.94,.93, and.97, respectively; P in all cases). Discrepancies (differences between PP and WHVP 5mm Hg) were observed in 7 patients (10%) (Fig. 1). In only 1 patient (with HCV related cirrhosis) the discrepancy was the result of an underestimation of portal pressure by WHVP. Thus, the prevalence of a presinusoidal component to the portal hypertension among HCV infected patients was 2% (or 1 out of 46). Six cases had a WHVP 5 mm Hg above PP. Two of them had HCV infection and 4 were alcoholics (Fig. 1). The clinical characteristics of these patients were similar to those observed in patients not having a WHVP above PP. Five of the 6 patients (83%) with a WHVP overestimating PP had 1 or more abnormalities in the portal blood perfusion of the liver: portal flow was hepatofugal in 3 patients, 3 patients had partial portal vein thrombosis, and in 1, most of the portal blood flow was diverted through a grossly dilated umbilical vein. These abnormalities, by contrast, were only observed in FIG. 1. Correlation between directly measured portal pressure (PP) and WHVP in all the patients studied according to the cause of liver disease. Values of Spearman (r) and intraclass correlation coefficient (Ri) for the overall series of patients and for each etiological group are also given.

4 1396 PERELLÓ ET AL. HEPATOLOGY December % of the patients without such a discrepancy between WHVP and PP (P.002). Agreement Between the Response of Portal Pressure and WHVP to Propranolol Administration. Nine patients had simultaneous measurements of WHVP and PP before and after the iv administration of propranolol. Propranolol caused a significant decrease in the portal pressure gradient (PPG) (from to mm Hg; P.01; mean change: mm Hg) and in HVPG (from to mm Hg; P.01; mean change: mm Hg). There was an excellent agreement between the PPG response to acute propranolol administration and the simultaneously measured HVPG response (r.84, Ri.87; P.004) (Fig. 2). DISCUSSION The present study investigated whether the measurement of WHVP and HVPG at hepatic vein catheterization, which is the preferred method to assess portal pressure in patients with chronic liver diseases, is adequate for that purpose in patients with chronic hepatitis and cirrhosis resulting from HCV infection, which has become the most common cause of chronic liver diseases in Western countries. 19 This investigation was prompted by the fact that earlier studies suggested that in nonalcoholic liver diseases, WHVP could frequently underestimate portal pressure because of a significant presinusoidal component in the portal hypertension. 11,12,16 If this would also apply to HCV related liver diseases, it would be necessary to substitute the safe and simple technique of hepatic vein catheterization for the direct puncture of the portal vein whenever an accurate measurement of portal pressure is required in this group of patients. This would add difficulties because this procedure is more invasive and demanding than the hepatic vein catheterization. Nonalcoholic liver disease is, however, an heterogeneous group, including PBC, chronic active autoimmune hepatitis, FIG. 2. Correlation between the absolute changes of portal pressure gradient (PPG) and HVPG after the acute iv administration of propranolol. Wilson s disease, hemochromatosis, and HBV and HCV related chronic liver diseases. We and others have confirmed that PBC is very frequently associated with a presinusoidal component when causing portal hypertension, 12,18 which makes the use of HVPG to estimate portal pressure in this condition inadequate. Chronic active autoimmune hepatitis and cirrhosis are also known to be frequently associated with a WHVP lower than PP. 11,12 Among the remaining nonalcoholic causes of cirrhosis and/or portal hypertension, chronic liver disease resulting from HCV and HBV infection is the most important, because of its very high prevalence. Although Lin et al. 17 have shown a good agreement between WHVP and PP in HBV related cirrhosis, no data are available regarding HCV related chronic liver diseases. Moreover, it should be noted that HCV infection is frequently associated with alcoholic liver disease, 19,20 and in these cases it is difficult to ascertain which of the two is the main responsible factor for the disease, which adds further uncertainty to the use of hepatic vein catheterization to monitor portal pressure whenever a patient has a positive test for HCV. The present study addressed this issue by performing simultaneous measurements of WHVP (by hepatic vein catheterization) and portal pressure (by direct percutaneous or transjugular puncture) in 3 groups of patients with chronic liver disease resulting from chronic HCV infection, alcohol abuse, and the association of both alcoholism and HCV infection. The most important finding of this study is that it showed an excellent agreement between the direct measurement of the portal venous pressure and the indirect technique of measuring WHVP at hepatic vein catheterization in the 3 groups of patients studied. Hepatic vein catheterization was as accurate to determine portal pressure in patients with chronic HCV infection as in alcoholics. Patients having both conditions: HCV infection and alcohol abuse, also showed an excellent concordance between WHVP and portal pressure. This agreement was found in patients with asymptomatic-toadvanced liver disease (Child-Pugh C cirrhosis). Moreover, it should be remarked that a discrepancy between both measurements of 5 or more mm Hg was observed in only 10% of the patients. There was only 1 case in which the WHVP underestimated the actual portal pressure; this was 1 of the 46 HCV infected patients included in the study, which represents a prevalence of the presinusoidal component to the portal hypertension of just 2%. Such a low incidence of discrepancies, as compared with some of the previous studies in nonalcoholics, 11,12,16 may be in part due to our use of the balloon occlusion technique rather than traditional wedging of the catheter. With the latter technique the territory of the liver included in the measurement is much smaller than when using the former, which increases the likelihood of discrepancies between WHVP and PP because of heterogeneity within the cirrhotic liver, i.e., of sampling error. The study further shows that the more common discrepancy between WHVP and PP was the overestimation of portal pressure by WHVP, which occurred in 6 cases: 4 were alcoholics and 2 had HCV infection. Most patients showing such an overestimation of portal pressure by WHVP had abnormalities in portal blood perfusion of the liver, including hepatofugal flow, grossly dilated umbilical vein, and/or partial portal vein thrombosis. The association of hepatofugal portal flow with a WHVP higher than portal pressure was previously described in isolated cases. Boyer et al. 12 reported 3 cases of

5 HEPATOLOGY Vol. 30, No. 6, 1999 PERELLÓ ET AL alcoholic patients with hepatofugal portal flow having the portal pressure overestimated by WHVP. Rector et al. 26 also observed a WHVP higher than portal pressure in 2 out of 3 patients showing a reversal of portal flow in a large series of 85 alcoholic patients. In this situation, the sinusoidal blood flow and the hepatic sinusoidal pressure are likely to be largely dependent on the hepatic arterial perfusion. In any case, our data suggest that the identification at Dopplerultrasound of hepatofugal portal flow, partial portal vein thrombosis, and/or of grossly dilated collaterals shall point to the possibility that WHVP exceeds PP, regardless of whether the patient is an alcoholic or an HCV infected. The second part of our study was aimed at investigating whether measuring the changes in WHVP and HVPG caused by propranolol adequately reflects the portal pressure response to propranolol in patients with HCV related or alcoholic cirrhosis. Although the studies evaluating the effects of vasoactive drugs on HVPG assume that similar changes will occur in PPG, previous studies regarding the concordance between both measurements have yielded conflicting results. Thus, Valla et al. 13 failed to show a good relationship between the response of WHVP to propranolol and that of portal pressure in 6 patients with alcoholic cirrhosis. In contrast, Rector et al. 14 reported a good agreement between both measurements following the administration of propranolol, also in 6 alcoholic cirrhotic patients. 14 Our results are consistent with the last study by showing an excellent agreement between the HVPG and the PPG response to propranolol in HCV and alcoholic cirrhosis. Although the number of patients studied was not much greater than in the previous studies, the consistency of the results provides strong support to the contention that changes in the PPG after -adrenergic blockade can be correctly evaluated by measuring changes in HVPG. In conclusion, the results of the current study corroborate that the simple and safe technique of hepatic vein catheterization, with measurements of WHVP, FHVP, and HVPG, is as adequate to assess portal venous pressure in HCV related liver disease as in alcoholic liver disease. Moreover, our results show that the measurement of HVPG accurately reflects the portal pressure response to propranolol administration, regardless of the cause of cirrhosis. This study further suggests that WHVP or HVPG measurements underestimating portal pressure in nonalcoholic cirrhosis is a possibility that has been overemphasized. 11,12,16 Probably the only disease where this is a real problem is PBC, 12,18 although in alcoholic or post-hepatitic cirrhosis this is a rare event, which occurs less frequently than the overestimation of portal pressure. Acknowledgment: The authors thank Ms. Diana Bird for her secretarial support and Angels Baringo, Laura Rocabert, and Rosa Sáez for expert technical assistance. REFERENCES 1. Groszmann RJ, Bosch J, Grace ND, Conn HO, García-Tsao G, Navasa M, Alberts J, et al. Hemodynamic events in a prospective randomized trial of propranolol vs placebo in the prevention of a first variceal hemorrhage. Gastroenterology 1990;99: Feu F, Garcia-Pagan JC, Bosch J, Luca A, Teres J, Escorsell A, Rodés J, et al. Relation between portal pressure response to pharmacotherapy and risk of recurrent variceal haemorrhage in patients with cirrhosis. Lancet 1995;346: Villanueva C, Balanzó J, Novella MT, Soriano G, Sainz S, Torras X, Cussó X, et al. Nadolol plus isosorbide mononitrate compared with sclerotherapy for the prevention of variceal rebleeding. NEJM 1996;334: Armonis A, Patch D, Burroughs AK. Hepatic Venous Pressure Measurement: An Old Test As a New Prognostic Marker in Cirrhosis? HEPATOL- OGY 1997;25: Ready JB, Robertson AD, Goff JS, Rector WG. Assesment of the risk of bleeding from esophageal varices by continuous monitoring of portal pressure. Gastroenterology 1991;100: Moitinho E, Escorsell A, Bandi JC, Salmerón JM, García-Pagán JC, Rodés J, Bosch J. Prognostic value of early measurements of portal pressure in acute variceal bleeding. Gastroenterology 1999;117: Merkel C, Bolognese M, Bellon S, Zuin R, Noventa F, Finuci G, Sacerdoti D, et al. Prognostic usefulness of hepatic vein catheterization in patients with cirrhosis and esophageal varices. Gastroenterology 1992;102: Tage-jensen U, Henriksen JH, Christensen E, Widding A, Ring-Larsen H, Juel N. Plasma catecholamine level and portal venous pressure as guides to prognosis in patients with cirrhosis. J Hepatol 1988;6: Vorobioff J, Groszmann RJ, Picabea E, Gamen M, Villavicencio R, Bordato J, Morel I, et al. Prognostic value of hepatic venous pressure gradient measurements in alcoholic cirrhosis. Gastroenterology 1996;111: Patch D, Armonis A, Sabin C, Christopoulou K, Greenslade L, McCormick A, Dick R, et al. Single portal pressure measurement predicts survival in cirrhotic patients with recent bleeding. Gut 1999;44: Pomier-Layrargues G, Kusielewicz D, Willems B, Villenueve JP Marleau D, Coté J, Huet PM. Presinusoidal portal hypertension in non-alcoholic cirrhosis. HEPATOLOGY 1985;5: Boyer TD, Triger DR, Horisawa M, Redeker AG, Reynolds TB. Direct transhepatic measurement of portal vein pressure using a thin needle. Comparison with wedged hepatic vein pressure. Gastroenterology 1977;72: Valla D, Bercoff E, Menu Y, Bataille C, Lebrec D. Discrepancy between wedged hepatic venous pressure and portal venous pressure after acute propranolol administration in patientes with alcoholic cirrhosis. Gastroenterology 1984;86: Rector WG. Propranolol for portal hypertension. Evaluation of therapeutic response by direct measurement of portal vein pressure. Arch Intern Med 1985;145: Bosch J, Mastai R, Kravetz D, Navasa M, Rodés J. Hemodynamic evaluation of the patient with portal hypertension. Sem Liver Dis 1986;6: Iwao T, Toyonaga A, Ikegami M, Sumino M, Oho K, Sakaki M, Shigemori H, et al. Wedged hepatic venous pressure reflects portal venous pressure during vasoactive drug administration in nonalcoholic cirrhosis. Dig Dis Sci 1994;39: Lin HC, Tsai YT, Lee FY, Chang TT, Wang SS, Lay CS, Lee SD, et al. Comparison between portal vein pressure and wedged hepatic vein pressure in hepatitis B-related cirrhosis. J Hepatol 1989;9: Navasa M, Parés A, Bruguera M, Caballería J, Bosch J, Rodés J. Portal hypertension in primary biliary cirrhosis. Relationship with histological features. J Hepatol 1987;5: EASL International Consensus Conference on Hepatitis C. Consensus statements. J Hepatol 1999;30: Parés A, Barrera JM, Caballería J, Ercilla G, Bruguera M, Caballería LL, Castillo R, et al. Hepatitis C virus antibodies in chronic alcoholic patients: association with severity of liver injury. HEPATOLOGY 1990;12: Zarski JP, Thelu MA, Moulin C, Rachail M, Seigneurin JM. Interest of the detection of hepatitis C virus RNA in patients with alcoholic liver disease. J Hepatol 1993;17: Pomier-Layrargues G, Villeneuve JP, Willems B, Huet PM. Systemic and hepatic hemodynamics after variceal hemorrhage: Effects of propranolol and placebo. Gastroenterology 1987;93: Groszmann R, Glickman M, Blei A, Storer E, Conn H. Wedged and free hepatic venous pressure measured with a balloon catheter. Gastroenterology 1979;76: Casado M, Bosch J, García-Pagán JC, Bru C, Bañares R, Bandi JC, Escorsell A, et al. Clinical events after transjugular intrahepatic portosystemic shunt: correlation with hemodynamic findings. Gastroenterology 1998;114: Deyo RA, Diehr PD, Patrick DL. Reproducibility and responsiveness of health status measures. Statistics and strategies for evaluation. Control Clin Trials 1991;12(Suppl): Rector WG, Hoefs JC Jr, Hossack KF, Everson GT. Hepatofugal portal flow in cirrhosis: observations on hepatic hemodynamics and the nature of the arterioportal communications. HEPATOLOGY 1988;8:16-20.