Altered Adrenergic Responsiveness of Endothelium-Denuded Hepatic Arteries and Portal Veins in Patients With Cirrhosis

Transcription

1 GASTROENTEROLOGY 1999;116: Altered Adrenergic Responsiveness of Endothelium-Denuded Hepatic Arteries and Portal Veins in Patients With Cirrhosis JÖRG HELLER,* MICHAEL SCHEPKE,* NINA GEHNEN,* GERHARD J. MOLDERINGS, ANDREAS MÜLLER, JOCHEN ERHARD, ULRICH SPENGLER,* and TILMAN SAUERBRUCH* Departments of *General Internal Medicine, Pharmacology, and Surgery, University of Bonn; and Department of Surgery, University of Essen, Germany Background & Aims: Patients with cirrhosis are characterized by a reduced splanchnic vascular resistance and a hyporeactivity to adrenergic vasoconstrictors. So far, their adrenergic splanchnic vascular responsiveness has not been evaluated in vitro. We compared responses to 1 - and 2 -adrenoceptor stimulation of hepatic arteries and portal veins of patients with cirrhosis undergoing transplantation with those of organ donors. Methods: Isometric contractions of endothelium-denuded vessel rings were induced cumulatively by methoxamine and relaxations by isoproterenol. Results are expressed as percentage of the contraction obtained by 85 mmol/l KCl or of the relaxation obtained by 100 mol/l papaverine, respectively. Results: Maximal methoxamine-induced contractions were reduced in cirrhotic hepatic arteries (cirrhosis, 51.8% 6.8%; donor, 89.9% 6.6%; P F 0.01) and portal veins (cirrhosis, 49.2% 6.4%; donor, 94.0% 5.3%; P F 0.01). In cirrhosis, isoproterenol induced a less marked relaxation of hepatic arteries (cirrhosis, 46.6% 3.2%; donor, 100.3% 4.4%; P F 0.01) but an increased relaxation of portal veins (cirrhosis, 41.9% 6.2%; donor, 26.2% 2.8%; P F 0.01). Conclusions: In cirrhosis, endothelium-free hepatic arteries are hyporeactive to 1 - and 2 -adrenoceptor agonists, and portal veins are hyporeactive to 1 - but hyperreactive to 2 -adrenoceptor agonists. These findings support the in vivo findings of a hyporesponsiveness to adrenergic vasoconstrictors in patients with cirrhosis. Patients with decompensated cirrhosis are characterized by a hyperkinetic circulation with low arterial pressure, reduced peripheral and splanchnic vascular resistance, activated sympathetic nervous system, 1 5 and hyporesponsiveness to adrenergic vasoconstrictors. 6 Studies investigating arterial vessels of animals with portal hypertension showed a hyporesponsiveness to adrenergic vasoconstrictors 7 11 as well as to several vasodilators. 12 This arterial hyporesponsiveness has been shown to persist after removal of the endothelium. 13 Studies investigating receptor-dependent contractions of the portal vein showed also a decreased adrenergic response in rats with partial portal vein ligation. 11,14 17 To date, there are only insufficient examinations on the adrenergic response of splanchnic vessels in humans, which may behave different to animals. 18 Thus, to assess the vascular smooth muscle adrenoceptor function, we investigated the response of both endothelium-denuded hepatic arteries and portal veins from patients with cirrhosis, obtained during liver transplantation, to vasoconstrictory 1 -adrenoceptor or vasodilatory 2 -adrenoceptor stimulation. These results were compared with those obtained in experiments with vessels of organ donors. Patients and Methods From 33 cirrhotic patients undergoing liver transplantation (15 women, 18 men; years; 9 Child Pugh A, 12 Child Pugh B, 12 Child Pugh C), 28 segments of the hepatic arteries and 17 segments of the portal veins were obtained. Twenty corresponding specimens of the hepatic arteries and 19 of the portal veins from 30 organ donors (10 women, 20 men; years) served as controls. The cause of cirrhosis was hepatitis B in 13 patients, alcohol in 7, hepatitis C in 4, autoimmune hepatitis in 3, primary sclerosing cholangitis in 2, primary biliary cirrhosis in 2, and secondary biliary cirrhosis in 2. We determined the cardiac output in 29 patients with cirrhosis before surgery and found it to be elevated in most of them (mean, 10.1 SEM 0.5 L/min). Death of the organ donors was caused by head trauma in 14, subarachnoidal hemorrhage in 12, suicidal cerebral hypoxy in 3, and oligodendroglioma in 1. Vessels from transplant patients who received corticosteroids, -adrenoceptor blockers, or other vasoactive medications before transplantation were excluded from the study. Preparation of the Vessels In the operating room, the vessels were immediately transferred into oxygenated (95% O 2, 5% CO 2 ) ice-cold modified Krebs-Henseleit solution (in mmol/l: NaCl, 118.3; Abbreviations used in this paper: L-NAME, N G -nitro-l-arginine methyl ester; UW solution, University of Wisconsin solution by the American Gastroenterological Association /99/$10.00

2 388 HELLER ET AL. GASTROENTEROLOGY Vol. 116, No. 2 KCl, 4.7; CaCl 2, 2.5; MgSO 4, 1.17; KH 2 PO 2, 1.18; NaHCO 3, 25.0; EDTA, 0.026; and glucose, 11.1). Vessels were cleaned of connective tissue and cut in 2-mm-wide rings. Because the aim of the study was to investigate adrenoceptor-induced vascular smooth muscle contractility, we had to exclude any influences of the endothelium. Therefore, the endothelium was removed mechanically by gently rubbing with a rough steel rod. Thereafter, the rings were suspended between 2 steel hooks in individual organ chambers filled with 20 ml modified Krebs Henseleit solution (Schuler Organbad, Hugo Sachs Elektronic, March, Germany). The solution was continuously bubbled with a mixture of 95% O 2 and 5% CO 2 and maintained at 37 C with an outer water jacket. One of the hooks was anchored to the organ chamber and the other was connected to an isometric force transducer (Fort 10; World Precision Instruments, Berlin, Germany). Transducer outputs were amplified (TBM4; World Precision Instruments) and recorded on a multichannel polygraph (Rikadenki R10, Tokyo, Japan). The vessels were allowed to equilibrate for 1 hour. A baseline tension of 2 g was adjusted. This tension was found in orientating experiments to be optimal for the induction of vasoconstriction of the hepatic artery and the portal vein. After submaximal precontraction with 40 mmol/l KCl, we added acetylcholine (10 5 mol/l) to confirm the absence of any intact endothelium. The organ chambers were then rinsed 3 times with warm Krebs Henseleit solution. Vascular Response to 1 -Adrenoceptor Stimulation Ten hepatic arteries and 9 portal veins of the cirrhotics were compared with 10 hepatic arteries and 10 portal veins of the organ donors. We used the selective 1 -adrenoceptor agonist methoxamine to assess the responsiveness of the blood vessels to 1 -adrenoceptor activation. However, at high concentrations methoxamine is able to stimulate 2 - and -adrenoceptors as well. Therefore, both 2 - and -adrenoceptors were routinely blocked with 10 6 mol/l rauwolscine and 10 5 mol/l propranolol, respectively. After proving the absence of intact endothelium as described above, we determined cumulative concentration-response curves for methoxamine (10 8 to 10 3 mol/l). In parallel, we carried out time-matched control experiments (administration of vehicle). Thereafter, maximum contraction of the vessel ring was induced by 85 mmol/l KCl. Vascular Response to 1 -Adrenoceptor Stimulation After Preincubation With the Nitric Oxide Synthase Inhibitor N G -Nitro-L-Arginine Methyl Ester Rings of the hepatic arteries and portal veins of 5 cirrhotics were prepared as described above. Pairs of rings from each vessel were simultaneously examined. After proving successful endothelial removal and blockade of - and 2 - adrenoceptors (as described above), one ring was incubated with N G -nitro-l-arginine methyl ester (L-NAME) (10 5 mol/l) for 30 minutes before and during the experiment, whereas the other ring served as control. Isometric vessel contraction was measured after stimulation with methoxamine (10 8 to 10 3 mol/l). Finally, maximum contraction of the blood vessel ring was induced by 85 mmol/l KCl. Vascular Response to 2 -Adrenoceptor Stimulation We compared 13 hepatic arteries and 8 portal veins from the patients with cirrhosis with 10 hepatic arteries and 10 portal veins from the organ donors. The absence of intact endothelium was proven as described above. After the equilibration period, -adrenoceptors were irreversibly blocked by phenoxybenzamine (10 5 mol/l), and the blood vessel rings were precontracted with 40 mmol/l KCl. Cumulative dose response curves were determined for isoproterenol (10 8 to 10 3 mol/l). In parallel, time-matched control experiments (administration of vehicle) were performed. Maximum relaxation of the preparation was then induced by 10 4 mol/l papaverine. Influence of University of Wisconsin Solution on Vascular Contractility Because donor vessels were preoperatively stored in ice-cold University of Wisconsin solution (UW solution), whereas the vessels of the recipient were directly mounted into Krebs Henseleit solution, we investigated the influence of storage in UW solution on vascular response to adrenoceptor stimulation. Four rings of pig mesenteric arteries were cut in two equal parts. One part was immediately transferred in the organ chamber, and methoxamine dose-response curves were determined as described. The other part was stored in ice-cold UW solution for 5 hours before beginning the experiment. Chemical Materials All substances were purchased from Sigma Chemical Co. (St. Louis, MO) except for rauwolscine, papaverine, and isoproterenol, which were purchased from ICN (Aurora, OH), and the UW solution from Du Pont Critical Care (Waukegab, IL). The substances were freshly prepared in distilled water each day. All substances were added to the organ bath in a volume of 100 µl. Statistical Analysis Vasoconstriction after methoxamine application is expressed as the percentage of the maximum vasoconstriction induced by 85 mmol/l KCl. The relaxant response to isoproterenol is expressed as the percentage of the maximum vasodilatation induced by 100 µmol/l papaverine. This relative expression of values was chosen to normalize vessels for possible differences in vessel diameters or smooth muscle mass that might affect absolute contractions and relaxations independently from adrenoceptor density or adrenoceptor-dependent pathways. A curve was fitted for the respective mean values for contraction and relaxation by nonlinear regression using the computer program Prism (Graph Pad Software Inc., San Diego, CA). pec 50 values (negative logarithm of the concentration producing a half maximum effect) were calculated from the

3 February 1999 ALTERED VASCULAR RESPONSE IN CIRRHOSIS 389 fitted curve. Statistical analysis was performed by two-way analysis of variance (ANOVA). Additionally, values for maximal methoxamine- and KCl-induced contractions and maximal isoproterenol- and papaverine-induced relaxations were calculated in absolute values and expressed in grams. Statistical analysis of the differences between maximal contractions or maximal relaxations of the blood vessels from patients with cirrhosis and organ donors, respectively, as well as the influence of UW solution on vascular reactivity, was performed with a Student t test for unpaired data. Results UW solution did not affect the maximum contraction of pig mesenteric arteries by 85 mmol/l KCl (without UW solution, g; after 5-hour incubation in UW solution, g; n 4), or by 10 3 mol/l methoxamine (without UW solution, 43.8% 7.2%; after 5-hour incubation in UW solution, 44.6% 6.2%; n 4). The KCl-induced maximum contraction of portal veins and hepatic arteries was not different in organ donors and transplant recipients (Table 1). The methoxamine-induced contractions were significantly reduced in the hepatic arteries (cirrhosis, 51.8% 6.8%; donor, 89.9% 6.6%; P 0.01; Figure 1) and portal veins (cirrhosis, 49.2% 6.4%; donor, 94.0% 5.3%; P 0.01; Figure 2) of patients with cirrhosis. The pec 50 of methoxamine was similar in both groups (hepatic artery: cirrhosis, ; donor, ; portal vein: cirrhosis, ; donor, ). Preincubation with 10 5 mol/l L-NAME did not alter the contractile response of the hepatic arteries or the portal veins of patients with cirrhosis to 10 3 mol/l methoxamine (Table 2). The maximal vasorelaxation induced by 10 4 mol/l papaverine was significantly increased in hepatic arteries of patients with cirrhosis compared with those of organ donors. The papaverine-induced relaxations of the portal veins tended to be increased in the patients with cirrhosis (Table 1). Isoproterenol induced a significantly smaller Figure 1. Cumulative concentration-response curves to methoxamine (10 8 to 10 3 mol/l) in isolated, endothelium-denuded hepatic artery rings of patients with cirrhosis ( ) and organ donors ( ). Contractions by methoxamine are expressed as the percentage of the contraction induced by 85 mmol/l KCl. Each point represents the mean SEM. In some cases, SEM is less than the size of the symbols. The curves to the data are significantly different (P ; two-way ANOVA). Significant differences of the corresponding observed contractile responses at the different methoxamine concentrations between patients with cirrhosis and organ donors are indicated by asterisks (*P 0.05, **P 0.01; Student t test for unpaired data). relaxation of the hepatic arteries from patients with cirrhosis than of those from organ donors (cirrhosis, 46.6% 3.2%; donor, 100.3% 4.4%; P 0.01; Figure 3). In contrast, isoproterenol-induced relaxations of the portal veins were significantly pronounced in portal veins of patients with cirrhosis than in those of organ donors (cirrhosis, 41.9% 6.2%; donor, 26.2% 2.4%; P 0.01; Figure 4). The pec 50 values of isoproterenol were significantly higher in arteries from Table 1. Drug-Induced Contraction and Relaxation of Hepatic Arteries and Portal Veins From Patients With Cirrhosis Undergoing Transplantation and From Organ Donors Hepatic artery Portal vein Cirrhosis Donor Cirrhosis Donor Vasocontraction (g ) by 85 mmol/l KCl Vasocontraction (g ) by 1 mmol/l methoxamine Precontraction (g ) by 40 mmol/l KCl Vasorelaxation (g ) by 0.1 mmol/l papaverine a a Vasorelaxation (g ) by 1 mmol/l isoproterenol b b NOTE. Maximal vasorelaxation was significantly increased in the hepatic artery of patients with cirrhosis. a P 0.005, b P 0.05; Student t test for unpaired data. Vasoconstriction to 85 mmol/l KCl and 1 mmol/l methoxamine (top). Vasorelaxation to 0.1 mmol/l papaverine and 1 mmol/l isoproterenol of blood vessel rings precontracted by 40 mmol/l KCl (bottom).

4 390 HELLER ET AL. GASTROENTEROLOGY Vol. 116, No. 2 Figure 2. Cumulative concentration-response curves to methoxamine (10 8 to 10 3 mol/l) in isolated, endothelium-denuded portal vein rings of patients with cirrhosis ( ) and organ donors ( ). Contractions by methoxamine are expressed as the percentage of the contraction induced by 85 mmol/l KCl. Each point represents the mean SEM. In some cases, SEM is less than the size of the symbols. The curves to the data are significantly different (P ; two-way ANOVA). Significant differences of the corresponding observed contractile responses at the different methoxamine concentrations between patients with cirrhosis and organ donors are indicated by asterisks (*P 0.05, **P 0.01; Student t test for unpaired data). patients with cirrhosis than in donor arteries (cirrhosis, ; donor, ; P 0.05) but were not significantly different in portal veins (cirrhosis, ; donor, ). Discussion In the present study, we found that endotheliumdenuded splanchnic vessels from patients with cirrhosis show an impaired contraction to 1 -adrenoceptor stimulation compared with vessels from noncirrhotic organ donors. Such an altered response was also observed for Table 2. Contractile Response of Endothelium-Denuded Hepatic Arteries and Portal Veins From Patients With Cirrhosis to 1 mmol/l Methoxamine and Interaction With 10 µmol/l L-NAME Hepatic artery Portal vein Absence of L-NAME Presence of L-NAME NOTE. L-NAME did not change the contractility of the endotheliumdenuded vessels of patients with cirrhosis. Student t test for unpaired data. Data expressed as percent of the maximal contraction induced by 85 mmol/l KCl. Figure 3. Cumulative concentration-response curves to isoproterenol (10 8 to 10 3 mol/l) in isolated, endothelium-denuded hepatic artery rings of patients with cirrhosis ( ) and organ donors ( ). Vessels were precontracted by 40 mmol/l KCl. Relaxations induced by isoproterenol are expressed as the percentage of the reduction in vascular tone induced by papaverine (100 µmol/l). Each point represents the mean SEM. In some cases, SEM is less than the size of the symbols. The curves to the data are significantly different (P ; two-way ANOVA). Significant differences of the corresponding observed relaxing responses at the different isoproterenol concentrations between patients with cirrhosis and organ donors are indicated by asterisks (*P 0.05, **P 0.01; Student t test for unpaired data). relaxation experiments with cirrhotic hepatic arteries, whereas portal veins showed an enhanced relaxation on 2 -adrenoceptor stimulation in comparison to the donors. Several investigators have suggested that a hyporesponsiveness of the systemic and splanchnic vasculature to endogenous vasoconstrictors may contribute to the hyperdynamic circulation observed in portal hypertension. 6,19 However, results obtained in animal models of cirrhosis or portal hypertension are conflicting. They showed hyporesponsiveness, 7 11 normal vascular response, 20 or even hyperresponsiveness 21 to 1 -adrenoceptor stimulation of arterial vascular tissue. Portal veins were hyporeactive to norepinephrine in rats with partial portal vein ligation, 14,15,17 but hyperreactive in cirrhotic rats. 15 Studies investigating -adrenoceptors in patients with cirrhosis found either a hyporesponsiveness of dorsal hand veins to phenylepinephrine, 22 or no differences in the density of thrombocytic 2 -adrenoceptors, 23 or increased chronotropic and vasopressor responses to phenyl-

5 February 1999 ALTERED VASCULAR RESPONSE IN CIRRHOSIS 391 Figure 4. Cumulative concentration-response curves to isoproterenol (10 8 to 10 3 mol/l) in isolated, endothelium-denuded portal vein rings of patients with cirrhosis ( ) and organ donors ( ). Vessels were precontracted by 40 mmol/l KCl. Relaxations induced by isoproterenol are expressed as the percentage of the reduction in vascular tone induced by papaverine (100 µmol/l). Each point represents the mean SEM. In some cases, SEM is less than the size of the symbols. The curves to the data are significantly different (P ; two-way ANOVA). Significant differences of the corresponding observed relaxing responses at the different isoproterenol concentrations between patients with cirrhosis and organ donors are indicated by asterisks (*P 0.05, **P 0.01; Student t test for unpaired data). epinephrine. 24 All these equivocal results may be due to differences in the animal models or to the use of different materials. Recently, it has been shown that the response of human intrahepatic resistance arteries to vasoactive compounds cannot be predicted from animal experiments. 18 Data on the vascular reactivity of patients with liver disease are scarce (for review see Hadoke and Hayes 25 ). Our study revealed a hyporesponsiveness to methoxamine of endothelium-denuded hepatic arteries and portal veins in patients with cirrhosis, which is in agreement with the report of Smith et al. 26 who found a hyporeactivity to phenylepinephrine of hepatic arteries from patients with hepatic failure. Our present data are in contrast to a study by Hadoke et al. published in abstract form, 27 in which no differences in phenylepinephrine- or norepinephrine-induced vasoconstriction between the hepatic arteries from patients with cirrhosis and organ donors were found. The reasons for this discrepancy may be a different study design, e.g., the use of different agonists, and the additional blockade of 2 - and -adrenoceptors performed in our experiments. To our knowledge, this is the first study presenting evidence for a hyporesponsiveness of the portal vein of patients with cirrhosis to 1 -adrenoceptor stimulation. This vascular hyporesponsiveness to 1 -adrenoceptor stimulation might be due to receptor down-regulation, to alterations at the postreceptor level of signal transduction, to defects of the contractile proteins, or to formation of counterbalancing vasodilating substances. Because the maximum vasoconstriction induced by receptor-independent membrane depolarization with KCl was not impaired, a defect in the contractile proteins is unlikely. Soluble vasodilatory substances (e.g., glucagon) should be washed out during the preparation of the vessels. UW solution has been shown to alter the acetylcholineinduced, endothelium-dependent vasorelaxation of human saphenous veins, but after storage at 4 C, neither norepinephrine-induced contraction nor sodium nitroprusside-induced relaxation was altered. 28 We proved, at least for pig mesenteric arteries, that storage in this solution did not affect the contractile response to methoxamine. Because in our experiments the endothelium had been removed from the vessels, a hyporesponse caused by endothelium-derived relaxing factors is highly improbable. Our results support observations reported by Michielsen et al., 13 who described a persistent hyporesponsiveness to 1 -adrenoceptor stimulation of the endotheliumfree thoracic aorta of portal vein ligated rats. Because in the present study - and 2 -adrenoceptors were blocked and a relatively selective 1 -adrenoceptor agonist was used, the observed reduced contractile response is most likely due to differences in 1 -adrenoceptor reactivity and not caused by a simultaneous -adrenoceptor induced relaxation, which could occur when using nonselective agonists such as norepinephrine. It has been suggested that an increased vascular production of nitric oxide induces the hypocontractility in portal hypertension. 29 However, in our experiments 10 5 mol/l L-NAME (an unspecific nitric oxide synthase inhibitor) did not change the reduced contractile response to methoxamine of the blood vessels from cirrhotic patients. By contrast, in the study performed by Smith et al., 26 hyporeactivity of recipient hepatic arteries was corrected by incubation with 10 4 mol/l L-NAME. In addition, inducible nitric oxide synthase messenger RNA was detected in the vessels. However, in this study patients with fulminant hepatic failure and patients with cirrhosis were mixed. It is not clear if inducible nitric oxide synthase was equally expressed in both groups. The concentration of 10 5 mol/l L-NAME might be too low to completely block inducible nitric oxide synthase. However, if nitric oxide overproduction would play an important role for vascular hyporesponsiveness in endothelium-denuded cirrhotic

6 392 HELLER ET AL. GASTROENTEROLOGY Vol. 116, No. 2 vessels, this concentration should at least partially have increased contractile response. Thus, according to our data, hyporesponsiveness to methoxamine in cirrhosis is more likely caused by down-regulation of 1 -adrenoceptors or by an alteration in the signal transduction cascade, such as postulated for portal-hypertensive rats. 10,30 Data on the -adrenoceptor density and function in portal hypertension are conflicting. -Adrenoceptor density on the blood lymphocytes from patients with cirrhosis 31 and from the cardiomyocyte plasma membranes of rats with cirrhosis 32 has been shown to be down-regulated. 2 -Adrenoceptor mediated relaxation was found to be impaired in the mesenteric veins from portal-hypertensive rats. 33 In contrast, -adrenoceptor density has been shown to be up-regulated in plasma membrane preparations from the portal veins of cirrhotic rats. 34 In the present study, an increased relaxation of the portal vein after -adrenoceptor stimulation was observed, whereas a reduced -adrenoceptor induced relaxation of the hepatic artery of patients with cirrhosis was found. This points to a dissociated regulation of the arterial and venous splanchnic vascular bed in these patients. Our results are compatible with findings in portal-hypertensive rabbits, in which a decreased G protein activated adenylate cyclase activity (which also mediates the -adrenoceptor induced vasodilatation) of the mesenteric artery, but an increased enzyme activity of the portal vein 35 was observed. The cause for this dissociated -adrenergic response in cirrhosis is unclear. The present study proves for the first time the existence of functional vasodilatory -adrenoceptors in the human portal vein. On the basis of in vitro data from human mesenteric veins 36 and in vivo data from animal studies, 37 it has been postulated that the human portal vein is devoid of -adrenoceptors. The present opposite finding suggests that -adrenoceptor blockers do not only act by reducing the cardiac output or the arterial smooth muscle tone but also by influencing the smooth muscle tone of the portal vein. Although the -adrenoceptor induced relaxation of the hepatic arteries of patients with cirrhosis was decreased compared with controls, the receptor-independent papaverine-induced relaxation of hepatic arteries and (not significantly) also of the portal vein was increased. It is conceivable that the enhanced receptorindependent relaxation of these blood vessels in patients with cirrhosis compared with healthy organ donors reflects a higher level of intrinsic vascular tone in the patients with cirrhosis, as shown for the isolated perfused cirrhotic rat liver. 38 In conclusion, the present study reveals an endothelium- and presumably nitric oxide independent hyporeactivity of hepatic arteries and portal veins to the 1 -adrenoceptor stimulation of patients with cirrhosis undergoing liver transplantation. The reactivity of hepatic arteries to -adrenoceptor stimulation was decreased, but the reactivity of portal veins to -adrenoceptor stimulation was increased. The decreased responses to 1 -adrenergic stimulation and the increased response of the portal vein to -adrenergic stimulation might explain an increased blood pooling in the territory of the prehepatic arterial and venous vessels as discussed for patients with cirrhosis. References 1. Flora JS, Legeault L, Morali GA, Hara K, Blendis LM. Increased sympathetic outflow in cirrhosis and ascites: direct evidence from intraneural recordings. Ann Intern Med 1991;114: Bichet DG, van Putten VJ, Schrier RW. Potential role of increased sympathetic activity in impaired sodium and water excretion in cirrhosis. N Engl J Med 1982;307: Henriksen JH, Ring-Larsen H, Christensen NJ. Sympathetic nervous activity in cirrhosis. J Hepatol 1984;1: Nicholls KM, Shapiro MD, Van Putten YJ, Kluge R, Chung HM, Bichet DG, Schrier RW. Elevated plasma norepinephrine concentrations in decompensated cirrhosis: association with increased secretion rates, normal clearance rates and suppressibility by central volume expansion. Circ Res 1985;56: Garcia-Pagan JC, Navasa M, Rivera F, Bosch J, Rodes J. Lymphocyte 2 -adrenoceptors and plasma catecholamines in patients with cirrhosis. Gastroenterology 1992;102: Lunzer MR, Manghani KK, Newman SP, Sherlock SPV, Bernard AG, Ginsburg J. Impaired cardiovascular responsiveness in liver disease. Lancet 1975;2: Polio J, Sieber CC, Lerner E, Groszmann RJ. Cardiovascular hyporesponsiveness to norepinephrine, propanolol, and nitroglycerin in portal-hypertensive and aged rats. Hepatology 1993;18: Bomzon A, Blendis LM. Vascular reactivity in experimental portal hypertension. Am J Physiol 1987;257:G158 G Sogni P, Sabry S, Moreau R, Gadano A, Lebrec D, Dinh-Xuan AT. Hyporeactivity of mesenteric resistance arteries in portal hypertensive rats. J Hepatol 1996;24: Huang YT, Wang GF, Yang MCM, Chang SP, Lin HC, Hong CY. Vascular hyporesponsiveness in aorta from portal hypertensive rats: possible sites of involvement. J Pharmacol Exp Ther 1996; 278: Liao JF, Yu PC, Lin HC, Lee FY, Kuo JS, Yang MCM. Study on the vascular reactivity and 1 -adrenoceptors of portal hypertensive rats. Br J Pharmacol 1994;111: Safka V, Moreau R, Gadano A, Lebrec D. Vascular hyporesponsiveness to vasodilators in rats with cirrhosis. J Hepatol 1997;26: Michielsen PP, Boeckxstaens GE, Sys SU, Herman AG, Pelckmans PA. The role of increased nitric oxide in the vascular hyporeactivity to noradrenaline in long-term portal vein ligated rats. J Hepatol 1995;23: Johansson B. Structural and functional changes in rat portal veins after experimental portal hypertension. Acta Physiol Scand 1976;98: Bomzon A, Jacob G, Lee SS, Meddings J. In vitro vascular responsiveness to norepinephrine in experimental portal hypertension. Clin Invest Med 1991;14: Jacob G, Bishara B, Lee SS, Hilzenart N, Bomzon A. Cardiovascular responses to serotonin in experimental liver disease. Hepatology 1991;14:

7 February 1999 ALTERED VASCULAR RESPONSE IN CIRRHOSIS Huang YT, Lin HC, Yu PC, Lee FY, Tsai YT, Lee SD, Yang MCM. Decreased vascular reactivity of portal veins in rats with portal hypertension. J Hepatol 1996;24: Battaglia SE, Jones RM, Angus JA, Smallwood RA, Angus PW. Control of the hepatic arterial circulation: humans are different (abstr). Hepatology 1997;26:275A. 19. Groszmann RJ. Hyperdynamic circulation of liver disease 40 years later: pathophysiology and clinical consequences. Hepatology 1994;20: Bomzon A, Binah O, Blendis LM. Hypotension in experimental cirrhosis: is loss of vascular responsiveness to norepinephrine the cause of hypotension in chronic bile duct ligated dogs? J Hepatol 1993;17: Cawley T, Geraghty J, Osborne H, Docherty JR. Effects of portal hypertension on responsiveness of rat mesenteric artery and aorta. Br J Pharmacol 1995;115: Bierbrier GS, Adams PC, Feldmann RD. Vascular -adrenergic responsiveness is reduced in cirrhosis. Clin Pharmacol Ther 1994;56: Mac Gilchrist AJ, Deighton NM, Hamilton CA, Reid JL. Binding studies of platelet 2 - and lymphocyte 2 -adrenoceptors in patients with cirrhosis. Br J Pharmacol 1990;30: Pinzani M, Marra A, Fusco BM, Alessandr M, Laffi G, Fanciullacci M, Gentillini P. Evidence for 1 -adrenoceptor hyperresponsiveness in hypotensive cirrhotic patients with ascites. Am J Gastroenterol 1991;86: Hadoke PWF, Hayes PC. In vitro evidence for vascular hyporesponsiveness in clinical and experimental cirrhosis. Pharmacol Ther 1997;75: Smith REA, Robinson NMK, McPeake JR, Baylis SA, Charles IG, Heaton ND, Moncada S, Williams R, Martin JF. Induction and role of NO synthase in hypotensive hepatic failure. Arterioscler Thromb Vasc Biol 1997;17: Hadoke PWF, Dillon JF, Walker SWW, Williams BC, Hayes PC. Hepatic arteries from cirrhotic patients are not hyporesponsive to vasoconstrictors (abstr). Hepatology 1996;24:145A. 28. Anastasiou N, Allen S, Paniagua R, Chester A, Yacoub M. Altered endothelial and smooth muscle cell reactivity caused by University of Wisconsin preservation solution in human saphenous vein. J Vasc Surg 1997;25: Vallance P, Moncada S. Hyperdynamic circulation in cirrhosis: a role for nitric oxide? Lancet 1991;337: Moreau R, Lebrec D. Endogenous factors involved in the control of arterial tone in cirrhosis. J Hepatol 1995;22: Gerbes AL, Remien J, Jungst D, Sauerbruch T, Paumgartner G. Evidence of down-regulation of 2 -adrenoceptors in cirrhotic patients with severe ascites. Lancet 1986;1: Lee S, Marty J, Mantz J, Samain E, Braillon A, Lebrec D. Desensitization of myocardial -adrenergic receptors in cirrhotic rats. Hepatology 1990;12: Martínez-Cuesta MA, Moreno L, Pique J, Bosch J, Rodrigo J, Esplugues JV. Nitric oxide mediated 2 -adrenoceptor relaxation is impaired in mesenteric veins from portal-hypertensive rats. Gastroenterology 1996;111: Gaudin C, Launay JM, Moreau R, Cailmail S, Lebrec D. Discrepant regulation among alpha and beta adrenoceptors in vascular walls of cirrhotic rats (abstr). J Hepatol 1994;21:S Cahill PA, Wu Y, Sitzmann JV. Altered adenylyl cyclase activities and G-protein abnormalities in portal hypertensive rabbits. J Clin Invest 1994;93: Tornebrandt K, Nobin A, Owman C. Pharmacological characterization of alpha-adrenergic receptor subtypes mediating contraction in human mesenteric arteries and veins. Blood Vessels 1985;22: Richardson PDI, Withrington PG. Alpha- and beta-adrenoceptors in the hepatic portal venous vascular bed of the dog. Br J Pharmacol 1978;62: Bhathal PS, Grossman HJ. Reduction of the increased portal vascular resistance of the isolated perfused cirrhotic rat liver by vasodilators. J Hepatol 1985;1: Received February 10, Accepted November 11, Address requests for reprints to: Jörg Heller, M.D., c/o Prof. Dr. Tilman Sauerbruch, Department of General Internal Medicine, University of Bonn, Sigmund-Freud-Strasse 25, D Bonn, Germany. Fax: (49) The authors thank Dr. H. Berthold, Dr. M. Flesch, and Prof. Dr. Böhm for excellent advice. Dedicated to Professor Dr. G. Strohmeyer on the occasion of his 70th birthday.