Randomized Comparison of Long-term Losartan Versus Propranolol in Lowering Portal Pressure in Cirrhosis

Transcription

1 GASTROENTEROLOGY 2001;121: LIVER, PANCREAS, AND BILIARY TRACT Randomized Comparison of Long-term Losartan Versus Propranolol in Lowering Portal Pressure in Cirrhosis JUAN GONZÁLEZ-ABRALDES,* AGUSTIN ALBILLOS, RAFAEL BAÑARES, LUIS RUIZ DEL ARBOL, EDUARDO MOITINHO,* CLAUDIO RODRÍGUEZ, MONICA GONZÁLEZ, ANGELS ESCORSELL,* JUAN CARLOS GARCÍA-PAGÁN,* and JAUME BOSCH* *Hepatic Hemodynamics Laboratory, Liver Unit, Institut de Malalties Digestives, Hospital Clinic, Institut de Investigacions Biomediques August Pi i Sunyer, Universitat de Barcelona, Barcelona; Gastroenterology Unit, Hospital Ramón y Cajal, Universidad de Alcala, Madrid; Hepatic Hemodynamics Laboratory, Liver Unit, Hospital General Universitario Gregorio Marañon, Universidad Complutense, Madrid; and Gastroenterology Unit, Hospital Puerta del Mar, Cádiz, Spain See editorial on page 487. Background & Aims: It has been suggested that losartan, an angiotensin II (A-II) type 1 receptor blocker, may have a pronounced portal pressure reducing effect, far greater than that of propranolol. This randomized controlled trial compared the hemodynamic and renal effects of continued 6-week administration of losartan (n 25) vs. propranolol (n 15) in portal hypertensive patients with cirrhosis treated endoscopically after a variceal bleeding episode. Methods: Hepatic venous pressure gradient (HVPG), systemic hemodynamics, renal function, and vasoactive factors were measured before and at 6 weeks of treatment. Results: Losartan did not reduce HVPG ( 2% 12%, NS) but significantly decreased mean arterial pressure (MAP, 8% 10%, P 0.001). On the contrary, propranolol significantly reduced HVPG ( 10% 11%, P 0.003) and cardiac output ( 16% 12%, P 0.001) but did not modify MAP (2.5% 10%, NS). Losartan increased A-II levels, reduced aldosterone, and decreased glomerular filtration rate (GFR) in Child B patients. Propranolol did not modify renal function. Adverse events related to therapy were mild and similar in both groups. Conclusions: Unlike propranolol, long-term losartan administration does not significantly reduce HVPG in patients with cirrhosis treated after a variceal bleeding episode, and it caused hypotension and reduced GFR in patients with moderate liver failure. Therefore, losartan is not an alternative to propranolol in preventing variceal rebleeding. Many studies have shown that nonselective betablockers, such as propranolol and nadolol, are effective in decreasing portal pressure and the risk of variceal hemorrhage. 1 When the hepatic venous pressure gradient (HVPG) is decreased to 12 mm Hg or below, there is an absolute protection from the risk of bleeding, a significant decrease in the size of varices, and an improvement in the actuarial probability of survival. 2 Even without reaching this target, it has been shown that reductions of the HVPG of more than 20% of baseline values are associated with an extremely low risk of variceal rebleeding. 3,4 Unfortunately such a satisfactory hemodynamic response is achieved in only about one third of the patients. 4 Also, approximately 15% of the patients may have contraindications to beta-blockers or will not tolerate this treatment. This has prompted research of alternative drugs for long-term treatment of portal hypertension. Recently, it has been suggested that the angiotensin II (A-II) type 1 receptor blocker losartan can achieve a marked reduction of HVPG in patients with cirrhosis, without significant adverse effects, 5 suggesting that losartan may be an alternative to beta-blockers in the long-term treatment of portal hypertension. However, previous experiences with other vasodilators 6,7 and with salarasin (a competitive nonselective A-II receptor blocker) 8 suggested that these treatments may cause significant arterial hypotension and adverse renal effects. 9 The present randomized controlled trial was performed to compare the effects of losartan with those of propranolol on splanchnic and systemic hemodynamics and renal function in patients with advanced cirrhosis. Abbreviations used in this paper: A-II, angiotensin II; ALD, aldosterone; CO, cardiac output; ET, endothelin; FHVP, free hepatic venous pressure; GFR, glomerular filtration rate; HVPG, hepatic venous pressure gradient; MAP, mean arterial pressure; RAS, renin-angiotensin system; SVR, systemic vascular resistance; WHVP, wedged hepatic venous pressure by the American Gastroenterological Association /01/$35.00 doi: /gast

2 August 2001 LOSARTAN VS. PROPRANOLOL IN PORTAL HYPERTENSION 383 Patients and Methods The study was conducted in patients with cirrhosis who had bled from esophageal varices, recruited from 4 university hospitals in Spain. The final protocol was approved by the Ethical Committee of each participating hospital and the Spanish Ministry of Health and Consumer Affairs. The study was conducted following the principles of the Declaration of Helsinki. Each patient gave written informed consent to participate in the study. Inclusion criteria were: (1) age between 18 and 75 years, (2) positive diagnosis of liver cirrhosis, and (3) variceal bleeding in the course of the previous year treated endoscopically with sclerotherapy or band ligation. The exclusion criteria were: (1) severe liver failure, evaluated by the presence of a serum bilirubin level 5 mg/dl, prothrombin rate 40%, and hepatic encephalopathy, (2) contraindications to beta-blockers (insulin-dependent diabetes mellitus, asthma, chronic obstructive lung diseases, atrioventricular block, heart rate 50 bpm, peripheral arterial disease), (3) serum creatinine 1.6 mg/dl, (4) hyperkaliemia (plasma potassium 5.5 meq/l), (5) systolic blood pressure 90 mm Hg, (6) hepatocellular carcinoma, (7) portal vein thrombosis, and (8) refusal to participate. Presence of ascites and/or treatment with diuretics were not considered exclusion criteria. Protocol Synopsis Patients fulfilling all inclusion criteria and no exclusion criteria were randomized to receive either propranolol or losartan. Losartan was a kind gift from Merck Sharp & Dohme, Spain. The dose was adjusted individually by a trained research nurse on an outpatient basis. For propranolol, starting dose was 20 mg twice daily, which was increased stepwise at 3-day intervals up to a maximum of 160 mg twice daily, if tolerated, as long as heart rate did not decrease below 55 bpm and/or systolic blood pressure below 90 mm Hg. For losartan, the initial dose was 6.25 mg/day, which was increased stepwise at 3-day intervals up to a maximum of 50 mg/day, if tolerated, as long as systolic blood pressure did not decrease below 90 mm Hg. Once the maintenance dose was reached, treatment was maintained for 6 weeks. Patients were seen at 3-week intervals by a physician inadvertent of the allocated treatment. Hemodynamic, renal, and vasoactive neurohumoral systems were evaluated before starting administration of the drug and at the end of the protocol. Sample Size Calculation and Randomization To test whether losartan may increase the number of patients showing a fall in HVPG of 20% of baseline values from the 35% maximal estimate with propranolol 4 to 90% (Schneider et al. reported 98% of patients achieving such a response with losartan 5 ), with alpha 0.05 and beta 0.10, a total of 14 patients in each arm was required. Because information on the effects of propranolol is available from many other published studies, whereas there is very scarce information on losartan, a biased randomization was chosen in a 3:5 ratio. This would provide 15 patients to be treated with propranolol and 25 with losartan. Randomization was computer generated in blocks of 8 patients. The code was kept in consecutively numbered, sealed, opaque envelopes. Every time a patient was introduced into the protocol, the coordinating center (Hospital Clinic) was contacted by fax to be informed of the allocated treatment. A total of 40 patients, 15 in the propranolol group and 25 in the losartan group, were finally included. Procedures Hemodynamic studies. Under fluoroscopic control, a 7 F balloon-tipped catheter (MediTech Cooper Scientific Corp., Watertown, MA) was advanced into the main right hepatic vein to measure wedged and free hepatic venous pressures (WHVP and FHVP, respectively). HVPG was calculated as the difference between WHVP and FHVP. A Swan-Ganz catheter (Abbott Laboratories, IL) was advanced into the pulmonary artery for measurements of cardiopulmonary pressures and cardiac output. A continuous thermal dilution catheter (Webster Laboratories, Inc., Baldwin Park, CA) was placed into the azygos vein for measurement of azygos blood flow (AzBF), according to previously described methods. 10 All measurements were performed at least in duplicate, and permanent tracings were obtained. Tracings were read blindly by the same observer (J.B.). Hepatic blood flow was estimated with the continuous infusion of indocyanine green, as previously described. 11 Mean arterial pressure (MAP) was measured noninvasively with an automatic sphygmomanometer, heart rate was derived from continuous electrocardiogram monitoring, and systemic vascular resistance (SVR) (dyn s 1 cm 5 ) was calculated as (MAP RAP)/CO 80, in which RAP (mm Hg) is the right atrial pressure and CO (L/min) is the cardiac output. Renal Function Renal function was evaluated by measurements of the creatinine clearance, blood urea nitrogen, 24-hour sodium excretion (U Na ), and glomerular filtration rate (GFR) evaluated as the clearance of radiolabeled ethylenediaminetetraacetic acid, as described previously. 7 Hormonal Studies Circulating A-II, aldosterone (ALD), norepinephrine, and endothelin (ET) were measured as previously described. 12,13 Nitrates and nitrites (NOx) were determined by the Griess method as described previously. 14 Statistical Analysis Data are shown in the text as mean SD. Paired Student t test was used to assess the significance of comparisons with baseline within each group and unpaired Student t test for comparisons between groups. Proportional data were analyzed by the Fisher exact test. Significance was established at

3 384 GONZÁLEZ ABRALDES ET AL. GASTROENTEROLOGY Vol. 121, No. 2 Table 1. Baseline Clinical Characteristics of the Patients P All statistics were computed using the SPSS statistical package 9.0 (SPSS, Chicago, IL). Results Losartan (n 23) Propranolol (n 14) P Age NS Sex (M/F) 20/3 11/3 NS Etiology (alcohol/hcv/hbv) 14/8/1 5/8/1 NS Child (A/B/C) 11/12/0 5/7/2 NS Child score (mean SD) NS History of ascites (%) NS Diuretic treatment (%) NS Variceal size 5 mm(%) NS Hematocrit at inclusion NS Time from index bleeding (days) NS HVPG at inclusion (mm Hg) NS Mean arterial pressure (mm Hg) NS Cardiac output (L/min) NS Creatinine clearance (ml/min) NS BUN (mg/dl) NS A-II (pg/ml) NS ALD (ng/dl) NS NE (pg/ml) NS NOx (nmol/ml) NS ET (pg/ml) NS Dose after titration, median (range) (mg/day) 50 ( ) 50 ( ) HCV, hepatitis C virus; HBV, hepatitis B virus; BUN, blood ureic nitrogen; NE, norepinephrine; NOx, nitrates and nitrites; NS, not significant. A total of 25 patients were randomized to receive losartan and 15 propranolol. Three patients did not complete the study because of side effects in 1 patient receiving losartan and after withdrawal of consent in 2 patients (1 propranolol, 1 losartan). Thus, the final analysis includes 23 patients receiving losartan and 14 propranolol. The main characteristics of these patients are summarized in Table 1. There were no differences between patients randomized to losartan or propranolol in any parameter. Hepatic Hemodynamics Propranolol caused a significant reduction of HVPG ( 10% 11%, P 0.003) (Table 2). In contrast, HVPG was not significantly modified after long-term therapy with losartan ( 2% 12%, NS) (Table 2 and Figure 1). HVPG was unchanged by losartan in 12 patients, increased in 5, and decreased in 6 patients by more than 10%. These 6 patients included a high proportion of alcoholics (5 of 6 vs. 9 of 17 in nonresponders, P 0.2) and had a higher baseline HVPG ( vs mm Hg, P 0.02). The 5 alcoholics had withdrawn from alcohol before entering into the study, 3 of them less than 3 months before randomization. In the global series, the effects of losartan on HVPG were similar in alcoholics and nonalcoholics (alcoholics: from to ; nonalcoholics: from to mm Hg; NS). Similarly, among patients treated with propranolol, there were no significant differences in the changes in HVPG in alcoholics vs. nonalcoholics ( to mm Hg, vs to mm Hg; NS). In both treatment groups, changes in HVPG were similar in patients receiving or not receiving diuretics (data not shown). Hepatic blood flow was not modified by propranolol or losartan. AzBF was significantly reduced by propranolol, but not by losartan. Systemic Hemodynamics Propranolol decreased heart rate and CO and increased PAP, which were not modified by losartan (Table 2). MAP was not modified by propranolol, but signifi- Table 2. Splanchnic and Systemic Hemodynamics Before and After 6-Week Treatment With Losartan or Propranolol Losartan Propranolol Baseline 6 wk P Baseline 6 wk HVPG (mm Hg) NS HBF (L/min) NS NS AzBF (L/min) NS PAP (mm Hg) NS PCWP (mm Hg) NS NS MAP (mm Hg) NS CO (L/min) NS NS SVR (day sec 1 cm 5 ) HR (bpm) NS HBF, hepatic blood flow; AzBF, azygos blood flow; PAP, pulmonary artery pressure; PCWP, pulmonary capillary wedged pressure; HR, heart rate; NS, not significant. P

4 August 2001 LOSARTAN VS. PROPRANOLOL IN PORTAL HYPERTENSION 385 Figure 1. Changes in (A) HVPG and (B) MAP during long-term losartan or propranolol administration. Results are expressed as mean SEM. cantly decreased under losartan (Figures 1 and 2). Hypotension was caused by a decrease in peripheral resistance and was not correlated with changes in HVPG. The hypotensive effect of losartan was related to the degree of liver failure. MAP did not decrease in Child A Figure 2. Comparison of the effects of long-term administration of losartan ( ) or propranolol ( ) on splanchnic and systemic hemodynamics. Results are expressed as percent change from baseline study (mean SEM). a Significantly different from baseline; b significantly different from losartan. patients ( 2% 8%, NS), whereas Child B patients showed a significant fall in MAP and SVR (Table 3). Renal Function and Vasoactive Systems Baseline renal function and vasoactive factors were similar in patients receiving propranolol or losartan (Tables 1 and 3). Propranolol caused no changes in GFR, blood urea nitrogen, creatinine, sodium excretion, and vasoactive factors (data not shown). Losartan caused an effective blockade of A-II receptors, reflected by an increase in circulating A-II levels and a decrease in plasma ALD, which was observed both in Child A and Child B patients (Table 3). Sodium excretion, norepinephrine, NOx, and ET were not significantly modified by losartan. Child B patients had higher baseline levels of A-II, ALD, and ET than Child A patients. The reduction in MAP and SVR observed in Child B patients after losartan was associated with a significant decrease in GFR (Table 3). Intake of diuretics was not related with changes in GFR ( 7% 26% vs. 5% 33% in patients taking diuretics vs. those not taking diuretics; NS). Adverse Events The overall incidence of adverse events was similar in the propranolol (27%) and losartan (28%) group (Table 4). However, serious adverse events were more

5 386 GONZÁLEZ ABRALDES ET AL. GASTROENTEROLOGY Vol. 121, No. 2 Table 3. Effects of Losartan in Patients of Child A and B Class Child A (n 11) Child B (n 12) Basal 6 wk P Basal 6 wk P HVPG NS NS MAP NS SVR NS GFR NS Cr NS NS BUN NS NS U Na NS NS A-II a c 0.05 ALD a NE NS c NS NOx NS NS ET b NS c NS U Na, urinary sodium excretion; BUN, blood ureic nitrogen; Cr, creatinine; NE, norepinephrine; NS, not significant. a P 0.10 vs. baseline value in Child B patients. b P 0.05 vs. baseline values in Child B patients. c P 0.05 vs. 6-week value in Child A patients. common under losartan, including the 2 instances of rebleeding during the study period. None of the adverse events were lethal, and only 1 resulted in withdrawal. Discussion Portal hypertension is the result of increased hepatic resistance and portal inflow. 15 Splanchnic vasoconstrictors reduce portal pressure by decreasing portal inflow and have been the mainstay of pharmacologic therapy. 1 Recently, the recognition of a dynamic component in the increased hepatic resistance of cirrhosis 16 has led to the introduction of vasodilators in the treatment of portal hypertension. 17 Endogenous factors, such as endothelin, A-II, and alpha-adrenergic stimulus increase hepatic vascular resistance, 18,19 whereas it is decreased by endogenous vasodilators like NO 20 or CO. 21 Vasodilators, however, may cause arterial hypotension, which is undesirable because it enhances the activation of endogenous vasoactive systems and water and sodium retention. 6,7 Table 4. Side Effects of Treatment Losartan Propranolol Symptomatic 1 0 hypotension Fatigue 2 3 Increase in the 1 1 diuretic dose Variceal bleeding 2 0 Others 1 (urinary lithiasis) 0 Total, n 7 (28%) 4 (27%) NS Withdrawals 2 (1 fatigue, 1 withdrawn consent) 1 (withdrawn consent) A recent study reported a dramatic reduction of portal pressure in cirrhotic patients treated with losartan (a nonpeptide antagonist of A-II receptor type 1), 5 which has focused attention on A-II as a target in the treatment of portal hypertension. 9 Increased A-II is the result of the activation of the renin-angiotensin system (RAS), which is commonly observed in patients with cirrhosis and has been shown to correlate with HVPG. 22 Further, A-II increases hepatic resistance in isolated cirrhotic livers 18,19 and infusion of A-II increases portal pressure and decreases hepatic blood flow in patients with cirrhosis. 23 These effects are probably caused by the contraction of vascular smooth muscle cells as well as hepatic stellate cells. 24 Activation of RAS may also worsen portal hypertension as a result of increased liver fibrogenesis, which may worsen the evolution of cirrhosis. All the above suggest that preventing the activation of RAS, or blocking the activity of A-II, may have beneficial effects in decreasing portal pressure in cirrhosis. However, stimulation of RAS is thought to represent a homeostatic response to peripheral vasodilatation. 28 Therefore, there is a risk that interfering with RAS may enhance peripheral vasodilatation and arterial hypotension, with adverse consequences on renal function. Actually, the A-II receptor antagonist salarasin caused marked arterial hypotension in patients with cirrhosis and ascites. 8 Similar effects were reported with the angiotensin-converting enzyme inhibitor captopril. 29 The present randomized controlled trial was specifically designed to assess whether long-term losartan treatment may favorably influence portal hypertension in patients with cirrhosis without adverse effects on systemic hemodynamics and renal function. The effects of

6 August 2001 LOSARTAN VS. PROPRANOLOL IN PORTAL HYPERTENSION 387 losartan were compared with those of propranolol because this is the accepted therapy for portal hypertension that any new drug should be tested against. 1 Unfortunately, the results of the current study show that losartan does not reduce portal pressure in patients with cirrhosis, but causes arterial hypotension. This hypotensive effect was caused by peripheral vasodilatation and was marked enough to be associated with a significant reduction in GFR in patients with moderately impaired liver function (Child B) and increased A-II levels. Losartan did not cause hypotension in Child A patients and had no beneficial effect in HVPG. In contrast, propranolol significantly decreased portal pressure and gastroesophageal collateral blood flow, without any negative effect on arterial pressure and renal function. These results argue strongly against any potential for losartan (and probably for any other A-II blocker) in the treatment of portal hypertension. This is supported by the preliminary results reported using irbesartan. 30 Our findings are in disagreement with those of Schneider et al. 5 who reported a marked fall in HVPG after losartan, with only a mild (but significant) decrease in MAP. Actually, their reported decrease in HVPG represents the greatest reduction in portal pressure ever described with any drug. An explanation for such discrepant results is not obvious. The dose of losartan used in our study was not insufficient because it was higher than in the study by Schneider et al. 5 ; this dose (median, 50 mg) was chosen following the standard schedule used in arterial hypertension, the main indication for losartan. 31 The fact that losartan increased circulating A-II and decreased ALD and arterial pressure indicates an effective blockade of angiotensin receptors. The reason why we were unable to confirm a beneficial effect of losartan on HVPG is unclear. The possibility of false-negative results is extremely unlikely because the study had enough power according to published studies. 4,5 The risk of selection bias was minimized by including a homogeneous group of patients, who had all experienced (and recovered from) a variceal bleeding episode and were under endoscopic therapy, and more importantly, by having a randomized allocation to therapy using state-of-the-art techniques in clinical trials. To prevent observer bias, doctors following the patients were unaware of whether the patients were taking propranolol or losartan. Moreover, the reading of all HVPG tracings was done blindly by an experienced observer. On the other hand, time to response cannot be the issue because in the study by Schneider et al., changes in HVPG were already present at 1 week. Moreover, when looking at individual responses in HVPG, only 6 of our patients showed a decrease in HVPG of more than 10% of baseline after losartan. It is important to note that 5 of these 6 patients were alcoholics, as were most of the patients in the study by Schneider et al. This may suggest that alcoholics are prone to respond to losartan; however, changes in HVPG were not different in alcoholics and nonalcoholics. An alternative explanation is that the fall in HVPG may have been caused by withdrawing from alcohol, 32 which was indeed observed in our patients. The probability of inadvertent selection and/or observer bias in the study by Schneider et al., thus, exists. This occurred in a previous study by the same group, reporting that spironolactone decreases HVPG by 10 mm Hg, 33 a finding that was not confirmed subsequently. 34,35 Our results may suggest that the physiologic role of A-II regulating intrahepatic resistance is of little significance. This is probably true, at least in advanced stages of cirrhosis. The possibility that in early cirrhosis, blockade of A-II might decrease hepatic resistance and portal pressure cannot be excluded. In summary, unlike propranolol, long-term losartan administration does not significantly reduce HVPG in cirrhotic patients treated after a variceal bleeding episode. In addition, losartan causes arterial hypotension and reduces the GFR in patients with moderate liver failure. Therefore, losartan should not be used as an alternative to propranolol in preventing variceal rebleeding. References 1. D Amico G, Pagliaro L, Bosch J. Pharmacological treatment of portal hypertension: an evidence-based approach. Semin Liver Dis 1999;19: Groszmann RJ, Bosch J, Grace ND, Conn HO, Garcia-Tsao G, Navasa M, Alberts J, Rodes J, Fischer R, Bermann M. Hemodynamic events in a prospective randomized trial of propranolol versus placebo in the prevention of a first variceal hemorrhage. Gastroenterology 1990;99: Merkel C, Bolognesi M, Sacerdoti D, Bombonato G, Bellini B, Bighin R, Gatta A. The hemodynamic response to medical treatment of portal hypertension as a predictor of clinical effectiveness in the primary prophylaxis of variceal bleeding in cirrhosis. Hepatology 2000;32: Feu F, Garcia-Pagan JC, Bosch J, Luca A, Teres J, Escorsell A, Rodes J. Relation between portal pressure response to pharmacotherapy and risk of recurrent variceal haemorrhage in patients with cirrhosis. Lancet 1995;346: Schneider AW, Friedrich J, Klein CP. Effect of losartan, an angiotensin II receptor antagonist, on portal pressure in cirrhosis. Hepatology 1999;29: Salmeron JM, Ruiz DA, Gines A, Garcia-Pagan JC, Gines P, Feu F, Claria J, Rivera F, Bosch J, Arroyo V. Renal effects of acute isosorbide-5-mononitrate administration in cirrhosis. Hepatology 1993;17: Albillos A, Lledo JL, Rossi I, Perez-Paramo M, Tabuenca MJ, Banares R, Iborra J, Garrido A, Escartin P, Bosch J. Continuous prazosin administration in cirrhotic patients: effects on portal

7 388 GONZÁLEZ ABRALDES ET AL. GASTROENTEROLOGY Vol. 121, No. 2 hemodynamics and on liver and renal function. Gastroenterology 1995;109: Arroyo V, Bosch J, Ribera F, Navarro-Lopez F, Rodes J. Effect of angiotensin-ii blockade on systemic and hepatic hemodynamics and on the renin-angiotensin-aldosterone system in cirrhosis with ascites. Eur J Clin Invest 1981;11: Garcia-Tsao G. Angiotensin II receptor antagonists in the pharmacological therapy of portal hypertension: a caution. Gastroenterology 1999;117: Bosch J, Groszmann RJ. Measurement of azygos venous blood flow by a continuous thermal dilution technique: an index of blood flow through gastroesophageal collaterals in cirrhosis. Hepatology 1984;4: Navasa M, Chesta J, Bosch J, Rodes J. Reduction of portal pressure by isosorbide-5-mononitrate in patients with cirrhosis. Effects on splanchnic and systemic hemodynamics and liver function. Gastroenterology 1989;96: Castro A, Jimenez W, Claria J, Ros J, Martinez JM, Bosch M, Arroyo V, Piulats J, Rivera F, Rodes J. Impaired responsiveness to angiotensin II in experimental cirrhosis: role of nitric oxide. Hepatology 1993;18: Guevara M, Gines P, Bandi JC, Gilabert R, Sort P, Jimenez W, Garcia-Pagan JC, Bosch J, Arroyo V, Rodes J. Transjugular intrahepatic portosystemic shunt in hepatorenal syndrome: effects on renal function and vasoactive systems. Hepatology 1998;28: Guarner C, Soriano G, Tomas A, Bulbena O, Novella MT, Balanzo J, Vilardell F, Mourelle M. Increased serum nitrite and nitrate levels in patients with cirrhosis: relationship to endotoxemia. Hepatology 1993;18: Bosch J, Garcia-Pagan JC. Complications of cirrhosis. I. Portal hypertension. J Hepatol 2000;32: Bathal PS, Grossmann HJ. Reduction of the increased portal vascular resistance of the isolated perfused cirrhotic rat liver by vasodilators. J Hepatol 1985;1: Garcia-Pagan JC, Escorsell A, Moitinho E, Bosch J. Influence of pharmacological agents on portal hemodynamics: basis for its use in the treatment of portal hypertension. Semin Liver Dis 1999;19: Rockey DC, Weisiger RA. Endothelin induced contractility of stellate cells from normal and cirrhotic rat liver: implications for regulation of portal pressure and resistance. Hepatology 1996; 24: Ballet F, Chretien Y, Rey C, Poupon R. Differential response of normal and cirrhotic liver to vasoactive agents. A study in the isolated perfused rat liver. J Pharmacol Exp Ther 1988;244: Wiest R, Groszmann RJ. Nitric oxide and portal hypertension: its role in the regulation of intrahepatic and splanchnic vascular resistance. Semin Liver Dis 1999;19: Suematsu M, Goda N, Sano T, Kashiwagi S, Egawa T, Shinoda Y, Ishimura Y. Carbon monoxide: an endogenous modulator of sinusoidal tone in the perfused rat liver. J Clin Invest 1995;96: Bosch J, Arroyo V, Betriu A, Mas A, Carrilho F, Rivera F, Navarro- Lopez F, Rodes J. Hepatic hemodynamics and the renin-angiotensin-aldosterone system in cirrhosis. Gastroenterology 1980; 78: Segel N, Bayley TJ, Paton A, et al. The effects of synthetic vasopressin and angiotensin on the circulation in cirrhosis of the liver. Clin Sci 1963;25: Bataller R, Gines P, Nicolas JM, Gorbig MN, Garcia-Ramallo E, Gasull X, Bosch J, Arroyo V, Rodes J. Angiotensin II induces contraction and proliferation of human hepatic stellate cells. Gastroenterology 2000;118: Rockey DC. Vasoactive agents in intrahepatic portal hypertension and fibrogenesis: implications for therapy. Gastroenterology 2000;118: McEwan PE, Gray GA, Sherry L, Webb DJ, Kenyon CJ. Differential effects of angiotensin II on cardiac cell proliferation and intramyocardial perivascular fibrosis in vivo. Circulation 1998;98: Border WA, Noble NA. Interactions of transforming growth factorbeta and angiotensin II in renal fibrosis. Hypertension 1998;31: Schrier RW, Arroyo V, Bernardi M, Epstein M, Henriksen JH, Rodes J. Peripheral arterial vasodilation hypothesis: a proposal for the initiation of renal sodium and water retention in cirrhosis. Hepatology 1988;8: Alexandre Pariente E, Bataille C, Bercoff E, Lebrec D. Acute effects of captopril on systemic and renal hemodynamics and on renal function in cirrhotic patients with ascites. Gastroenterology 1985;88: Schepke M, Werner E, Bieker E, Schiedermaier P, Hofer U, Layer G, Heller J, Neef M, Sauerbruch T. The angiotensin II receptor antagonist irbesartan for the treatment of portal hypertension: a placebo controlled, double blind study (abstr). Hepatology 2000; 32:406A. 31. Johnston CI. Angiotensin receptor antagonists: focus on losartan. Lancet 1995;346: Vorobioff J, Groszmann RJ, Picabea E, Gamen M, Villavicencio R, Bordato J, Morel I, Audano M, Tanno H, Lerner E, Passamonti M. Prognostic value of hepatic venous pressure gradient measurements in alcoholic cirrhosis: a 10-year prospective study. Gastroenterology 1996;111: Klein CP. Spironolactone in the treatment of portal hypertension in liver cirrhosis. A new therapeutic principle? Dtsch Med Wochenschr 1985;110: Garcia-Pagan JC, Salmeron JM, Feu F, Luca A, Gines P, Pizcueta P, Claria J, Piera C, Arroyo V, Bosch J. Effects of low-sodium diet and spironolactone on portal pressure in patients with compensated cirrhosis. Hepatology 1994;19: Nevens F, Lijnen P, VanBilloen H, Fevery J. The effect of long-term treatment with spironolactone on variceal pressure in patients with portal hypertension without ascites. Hepatology 1996;23: Received February 6, Accepted April 11, Address requests for reprints to: Jaume Bosch, M.D., Hepatic Hemodynamic Laboratory, Liver Unit, Hospital Clinic, Villarroel 170, Barcelona, Spain. jbosch@medicina.ub.es; fax: (34) Supported in part by grants from Plan Nacional de Investigación y Desarrollo (SAF and SAF ) and Fondo de Investigaciones Sanitarias (FIS 00/0444); and an award from the Hospital Clinic, Barcelona (to J. G.-A.).