Portal Vein-ligated Rats Research Articles

The role of inducible nitric oxide synthase in vascular hyporeactivity of endotoxin-treated and portal hypertensive rats

The involvement of the inducible nitric oxide (NO) synthase in the vascular hyporeactivity in portal vein-ligated rats was assessed in isolated perfused mesenteric arterial beds. Aminoguanidine, a selective inhibitor of the inducible NO synthase, restored the pressor responses to methoxamine in arteries of endotoxin-treated rats, but was ineffective in hyporeactive portal vein-ligated vessels. N G-Nitro- l-arginine methyl ester enhanced the responsiveness both in portal vein-ligated and sham-operated rats, without changing the difference between the two groups. These results not only indicate that the inducible NO synthase is not involved in the hyporeactivity to methoxamine in mesenteric arteries of portal hypertensive rats, but also suggest a role for factors other than NO.

Administration of Nω-nitro- l-arginine ameliorates portal-systemic shunting in portal-hypertensive rats

Background : Nitric oxide, a vasodilator synthesized from l-arglnine by vascular endothelial cells, may play a role in the development of portal-systemic collaterals. This study investigated the effect of long-term inhibition of NO secretion on portal systemic shunting. Methods : Systemic and splanchnic hemodynamics and the degree of portal-systemic shunting were evaluated in partial portal vein-ligated rats after administration of placebo (0.9% saline) or N ω-nitro- l-arginine (NNA) (~2 μg · kg −1 · min −1) intravenously for 6 days. Results : NNA treatment induced increases in splanchnic arterial resistance ( P < 0.001) and portal-collateral resistance ( P < 0.05) and a decrease in portal venous inflow ( P < 0.05). Portal pressure was not changed (NS). The splenic-systemic shunting was significantly decreased from 81% ± 5% in the placebotreated group to 69% ± 4% in the NNA-treated group ( P < 0.05), paralleled by an insignificant reduction in the mesenteric-systemic shunting (64% ± 7% vs. 50% ± 6%, NS). The attenuation of portal-systemic shunting by NNA was further shown by an increase in the vascular resistance of portal-systemic collateral venous bed using an in situ portal-systemic collateral perfusion model (1.27 ± 0.05 vs. 1.07 ± 0.03 cm H 2O · mL −1 · min −1; P < 0.001). Conclusions : The results show that in portal hypertensive rats, NNA reduces portal-systemic shunting without reducing portal pressure, suggesting that NO plays a role in the collaterallzation of the portal system. In addition, high flow through the portal-collateral bed is probably an important driving force that is independent of portal hypertension for the development of portal-systemic shunting in portal-hypertensive rats.

Endogenous nitric oxide (EDNO) production plasma volume and splanchnic blood flow following portal vein-ligation in rats

Endogenous nitric oxide (EDNO) production plasma volume and splanchnic blood flow following portal vein-ligation in rats

Studies on the safety of intrasplenic hepatocyte transplantation: relevance to ex vivo gene therapy and liver repopulation in acute hepatic failure.

Hepatocytes transplanted into the host liver engraft promptly, retain normal function, and survive indefinitely. Although intrasplenic transplantation is effective in delivering hepatocytes to the liver, to define potentially limiting complications, we studied its safety in normal, cirrhotic, and partial portal vein-ligated rats. In normal rats, portal pressures increased severalfold after hepatocyte transplantation but returned to normal within 3 weeks. In contrast, in portal hypertensive rats with partial portal vein ligation or cirrhosis, portal pressures were either unchanged or increased less after hepatocyte transplantation. However, more transplanted cells migrated to the lungs along with a rise in right atrial pressures in portal hypertensive rats. Further quantitative studies using 111Indium-labeled hepatocytes showed that intrasplenic retention of transplanted hepatocytes was similar in all animal groups. Intrahepatic cell translocation was comparable in normal and cirrhotic rats, whereas fewer cells migrated to the liver in partial portal vein-ligated rats. The most remarkable difference, however, was significantly greater intrapulmonary translocation of hepatocytes in portal hypertensive rats, which was presumably related to portosystemic shunting. These results indicate that because intrasplenic hepatocyte transplantation induces only temporary portal hypertension in normal subjects, potential strategies to augment liver repopulation could include repeated cell transplantation. This should be useful for optimizing the results of ex vivo gene therapy, or other hepatocyte-based therapies. However, the hepatic and portal hemodynamic status requires careful evaluation in portal hypertensive or cirrhotic subjects if serious complications are to be avoided.

The effect of portal hypertension on gastric epithelial proliferation in rats.

We investigated the effect of portal hypertension over time on gastric epithelial proliferation in rats. Portal-vein ligation (PVL) was applied to fourteen Wistar rats to produce experimental portal hypertension: seven rats (PVL2) had histological analysis after 2 months and the remaining seven (PVL4) after 4. An additional group of seven was employed as a control. Each rat was injected with bromodeoxyuridine and killed an hour later to assess gastric epithelial proliferation by an immunohistochemical staining technique. The labeling index between PVL and control animals differed significantly (p < 0.05). Labeling indices in fundic mucosa were 5.7 +/- 0.2% in controls, 3.9 +/- 0.3% in PVL2 rats and 4.0 +/- 0.3% in PVL4 rats, while in antral mucosa the corresponding figures were 7.5 +/- 0.6, 4.6 +/- 0.3 and 4.0 +/- 0.3%. In neither fundic nor antral mucosa was the total mucosal thickness significantly different between the control and PVL rats. In the PVL2 group, the frequency distribution of labeled cells showed a shift toward the mucosal surface in fundic glands. No significant differences between the two PVL groups were observed in any of the studied parameters. This study suggests that portal hypertension may inhibit gastric epithelial proliferation in rats.

Sodium retention and hepatic function following partial portal vein ligation in the rat

Recent studies have suggested that the development of sodium retention in experimental cirrhosis in the rat occurs when hepatic function, measured by the aminopyrine breath test, decreases below a critical threshold. The present study evaluated the relationship between renal sodium handling and hepatic function and determined whether sodium retention occurs following partial portal vein ligation. Sodium balance, urine volume, creatinine clearance and the aminopyrine rate constant of elimination, on a constant sodium intake, were evaluated daily, from 1 day before surgery to 5 days after surgery, in both sham-operated (n = 6) and partially portal vein-ligated rats (n = 14). In the partially portal vein-ligated group, sodium retention occurred in 9 rats between 1 and 4 days after surgery, accompanied by a 45% reduction in the aminopyrine rate constant of elimination. Spontaneous natriuresis occurred within 5 days after surgery, and was associated with an increase in the aminopyrine rate constant of elimination from 0.94 +/- 0.07 x 10(-2) min-1 on the last day of sodium retention to 1.36 +/- 0.06 x 10(-2) min-1 on the day of diuresis (P < 0.05). In contrast, creatinine clearance did not change throughout the study. There was a negative curvilinear association between sodium balance and the aminopyrine rate constant of elimination (r = 0.70, P < 0.001). In the five rats without sodium retention, there was no change in the aminopyrine rate constant of elimination and creatinine clearance over the 5 days.(ABSTRACT TRUNCATED AT 250 WORDS)

The role of nitric oxide in the vascular hyporesponsiveness to methoxamine in portal hypertensive rats.

This study examined whether an increased activity of the endothelium-derived relaxing factor, nitric oxide, may account for the hyporesponsiveness to vasoconstrictors in portal hypertension. We performed dose-response curves to methoxamine, an alpha-adrenoceptor agonist, with and without N omega-nitro-L-arginine, a specific inhibitor of nitric oxide synthesis, in experimental portal hypertension. Partial portal vein-ligated or sham-operated rats were pretreated with a continuous intravenous infusion of either N omega-nitro-L-arginine (50 micrograms.kg-1.min-1) or saline. Thirty minutes after starting the infusion of N omega-nitro-L-arginine or saline an infusion of methoxamine (10, 30 and 100 micrograms.kg-1.min-1) was added. Total peripheral resistance was calculated from mean arterial pressure and cardiac index. Repeated measurements of cardiac index were performed by a thermodilution technique. In portal vein-ligated rats pretreated with saline, the increase in total peripheral resistance after methoxamine infusion was significantly less than that of sham-operated rats (0.2 +/- 0.1 vs. 1.0 +/- 0.3, 0.6 +/- 0.1 vs. 1.6 +/- 0.3 and 3.7 +/- 0.5 vs. 6.1 +/- 0.7 mm Hg.ml-1.min.100 gm, p less than 0.05, methoxamine 10, 30 and 100 micrograms.kg-1.min-1, respectively). In the presence of N omega-nitro-L-arginine, the change in total peripheral resistance after methoxamine infusion was similar in both groups (p greater than 0.05). In conclusion, this study demonstrates that a vascular hyporesponsiveness to methoxamine is present in portal vein-ligated rats and that this hyporesponsiveness is reversed by blockade of nitric oxide.(ABSTRACT TRUNCATED AT 250 WORDS)

Portal-systemic shunting and the disruption of circadian locomotor activity in the rat

To determine if the extent of portal-systemic shunting (PSS) influences the disruption of circadian function in chronic liver disease, locomotor activity was examined in two rat models with varying degrees of PSS, i.e., portal vein ligation (PVL) and end-to-side portacaval anastomosis (PCA). Animals were housed in individual activity cages under conditions of 12 hour light/12 hour darkness (weeks 0–3), then under conditions of constant dim light (weeks 4–7). Cages were equipped with running wheels connected to a continuous recorder, and daily tracings of running activity were recorded for 7 weeks. Computer analysis of wheel revolutions per hour with a χ2 periodogram was used to calculate Qp, a measure of the amplitude of a circadian rhythm. The degree of PSS was measured by means of radioactive microspheres injected into the ileocolic vein and spleen. PVL rats were found to have PSS from the splenic and mesenteric territories of 88% and 27%, respectively; circadian periodicity was maintained in all PVL rats. PCA rats had complete shunting (>99%) and showed a range of disrupted circadian rhythms from blunting of the amplitude to complete absence of the locomotor activity rhythm. This spectrum of disorganization occurred in spite of similar degrees of liver atrophy and weight gain. Whereas PCA in rats markedly disturbs the circadian rhythm of locomotor activity, animals with considerably less PSS from PVL exhibit normal behavior. The extent of PSS could be a variable affecting the expression of circadian rhythms in liver disease.

Sequence of morphological and hemodynamic changes of gastric microvessels in portal hypertension

Changes in gastric microvasculature and blood flow at different phases of portal hypertension were studied in rats 1, 2, 3, 4, and 15 days after induction of portal hypertension or sham operation. Vessel lumen and vessel wall thickness were expressed as a ratio referred to the vessel size. On day 2 after constriction of the portal vein, gastric blood flow was decreased (0.57 ± 0.06 vs. 0.99 ± 0.20 mL · min−1 · 100 g−1; P < 0.05), and gastric vessels had a distended lumen (0.42 ± 0.02 vs. 0.28 ± 0.03; P < 0.01) and a thin wall (2.11 ± 0.2 vs. 3.82 ± 0.4; P < 0.01). On day 4, the gastric blood flow of portal hypertensive animals was increased (1.15 ± 0.14 vs. 0.71 ± 0.07 mL · min−1 · 100 g−1; P < 0.05), whereas gastric vessels had a reduced lumen (0.27 ± 0.02 vs. 0.33 ± 0.02; P < 0.01) and a thick wall (4.19 ± 0.52 vs. 3.16 ± 0.30; P < 0.05). By day 15, vessels with the largest lumens (0.45 ± 0.01 vs. 0.29 ± 0.01; P < 0.01) and the thinnest walls (1.78 ± 0.26 vs. 3.58 ± 0.62; P < 0.01) were observed in portal hypertensive animals. In conclusion, the gastric vessels of the 15-day portal vein-ligated rat resemble the structural abnormalities described in human portal hypertensive gastropathy.

Atrial natriuretic peptide in portal vein-ligated rats: alterations in cardiac production, plasma level and glomerular receptor density and affinity.

The atrial natriuretic peptide hormonal system is altered to a variable degree in patients with cirrhosis. Portal pressure and portal-systemic shunting are also varied in cirrhosis. We used a portal vein-ligated rat model with predictable portal hypertension to study the effects of portal hypertension alone on the atrial natriuretic peptide hormonal system. Sham-operated rats were used as controls. Mean portal pressure was significantly increased in portal vein-ligated rats (portal vein-ligated rats, 21.7 +/- 0.74 cm H2O; sham-operated rats, 13.7 +/- 0.47 cm H2O; p less than 0.0001). Plasma atrial natriuretic peptide decreased 50% in the portal vein-ligated rats (p less than 0.0001). Atrial natriuretic peptide messenger RNA level was decreased by 40% to 60% in the left and right atria and in the ventricles of portal vein-ligated rats (p less than 0.05 for each chamber). Only one class of glomerular binding site was identified by competitive binding studies. The atrial natriuretic peptide glomerular receptor density increased in the portal vein-ligated rats (portal vein-ligated rats, 1,660 +/- 393; sham-operated 725 +/- 147 fmol/mg protein, p less than 0.02), whereas affinity decreased (portal vein-ligated, 1.69 +/- 0.49; sham-operated, 0.55 +/- 0.12 nmol/L, p less than 0.02). No difference was seen in the amount of cyclic GMP generated by atrial natriuretic peptide stimulation in isolated glomeruli from portal vein-ligated and sham-operated rats.(ABSTRACT TRUNCATED AT 250 WORDS)

Portal-vein obstruction in children leads to growth retardation

The portal vein is the main source of blood and hepatotrophic factors to the liver. Partial portal-vein ligation in rats results in reduced growth compared with that in control rats. To investigate whether extrahepatic portal vein obstruction occurring in early childhood influences growth in humans, anthropometric and nutritional assessments were prospectively carried out in 61 patients with extrahepatic portal vein obstruction. Comparisons were made with 183 matched healthy controls using National Center of Health Statistics reference. Fifty-one percent of children with extrahepatic portal vein obstruction had stunted growth (height for age less than 90% of normal), compared with 16% of controls (p less than 0.01). Growth retardation was severe in patients with longer (greater than 5 yr) than with shorter (less than 2.5 yr) duration of clinical portal hypertension (height for age, 88.0 +/- 3.2 vs. 95.1 +/- 3.0; p less than 0.01). Little difference was seen in the energy intake (1,302 +/- 463 kcal/day vs. 1,335 +/- 449 kcal/day; p = not significant) and weight for height index (83.6 +/- 9.3 vs. 88.0 +/- 7.9; p = not significant) between extrahepatic portal vein obstruction patients and controls. This suggested that despite comparable nutrition, marked growth retardation occurred in extrahepatic portal vein obstruction patients. Incremental growth velocity was studied in 31 patients; in 24 (73%) the baseline Z score (-2.1 +/- 0.2) had decreased further (-2.4 +/- 0.2) at the end of follow-up (15.5 +/- 1.6 mo). Although the incremental height velocity was only 56% of the expected height, incremental weight gain was 98% of the expected weight for the attained height.(ABSTRACT TRUNCATED AT 250 WORDS)

Cardiovascular responses to serotonin in experimental liver disease

Recent evidence suggests that serotonergic mechanisms within the cardiovascular system are activated and may be important in the development of the hyperkinetic circulation and the maintenance of portal hypertension in cirrhotic patients. The in vivo pressor and positive chronotropic response together with the in vitro contractile responses of aortic rings and portal veins to serotonin were studied in three different rat models of cirrhosis, portal hypertension and jaundice: the portal vein-ligated rat, the carbon tetrachloride-induced cirrhotic rat, the chronic bile duct-ligated cirrhotic rat and the 3-day noncirrhotic, nonportal hypertensive-jaundiced rat. In addition, the activity of the enzyme monoamine oxidase type A was determined in lung homogenates prepared from the four groups of sham and treated animals. In the four different groups of sham-treated or operated pithed rats, serotonin caused a dose-dependent increase in mean arterial blood pressure without any effect on the heart rate. The pressor responses to serotonin in the three models of portal hypertension were significantly attenuated from their respective sham group. In the 3-day noncirrhotic, nonportal hypertensive-jaundiced rats, the pressor response was no different than that seen in the sham-operated rats. No evidence of consistent potentiated or blunted in vitro reactivity to serotonin of arterial rings and portal veins from the four groups of rats was seen. Portal hypertension, cirrhosis and hyperbilirubinemia had no effect on the activity of monoamine oxidase type A. These data demonstrate that portal hypertension is associated with an attenuated pressor response to serotonin.(ABSTRACT TRUNCATED AT 250 WORDS)

Endotoxin and the hyperdynamic circulation of portal vein—ligated rats

Humoral factors may be responsible for the hyperdynamic circulation seen in portal hypertension. Endotoxin, a peripheral arteriolar vasodilator, has been proposed to mediate this hemodynamic picture. We examined the pathogenic role of endotoxin in portal vein-ligated rats, a prehepatic portal hypertensive model with a well-developed hyperdynamic circulation. To this end, we (a) administered oral neomycin, a poorly absorbable antibiotic, at doses of 50 and 100 mg/day for 7 days and found no evident splanchnic hemodynamic effects of a 2-log-fold reduction of cecal aerobic bacterial flora as assessed by the radioactive microsphere technique in portal vein-ligated rats studied in the postanesthesia awake state; (b) assayed endotoxin in arterial samples using a quantitative limulus assay and found no evidence of endotoxinemia in PVL rats; (c) induced a state of endotoxin tolerance by repeated daily intraperitoneal injections of low-dose endotoxin and found no amelioration of the hyperdynamic state in portal vein-ligated rats. Our results do not support the hypothesis that endotoxin plays a major pathogenic role in the hyperdynamic circulation of this experimental model.

Effect of ethanol on splanchnic hemodynamics in awake and unrestrained rats with portal hypertension.

Alcoholic liver disease is frequently accompanied by portal hypertension. We have previously shown that alcohol intake in awake, unrestrained rats is followed by an increase in portal tributary blood flow. In this study, the effect of ethanol on splanchnic hemodynamics in rats with portal hypertension was analyzed. Portal hypertension was induced by partial ligation of the portal vein. This procedure resulted in an increase in portal tributary and hepatic arterial blood flows compared to sham-operated animals. Ethanol (2 gm per kg, oral) increased portal tributary blood flow in both sham-operated and portal vein-ligated rats (sham + water = 37.6 +/- 1.4; sham + ethanol = 63.1 +/- 1.9; p less than 0.01; partial portal vein stenosis + water = 53.2 +/- 3.3; partial portal vein stenosis + ethanol = 69.5 +/- 2.2 ml.kg-1.min-1; p less than 0.01). In sham-operated rats, hepatic artery blood flow was unchanged following ethanol (sham + water = 6.6 +/- 0.7; sham + ethanol = 7.1 +/- 1.0 ml.kg-1.min-1), whereas in portal vein-ligated rats, flow was increased (partial portal vein stenosis + water = 13.7 +/- 1.4; partial portal vein stenosis + ethanol = 19.8 +/- 1.1 ml.kg-1.min-1; p less than 0.025). The adenosine receptor blocker 8-phenyltheophylline suppressed only the ethanol-induced increase in both portal tributary and hepatic artery blood flows in portal vein-ligated rats. The increases in hepatic artery and portal tributary blood flows observed in portal vein-ligated rats without ethanol were not influenced by 8-phenyltheophylline.(ABSTRACT TRUNCATED AT 250 WORDS)

Portal venous pressure and portasystemic shunting in experimental portal hypertension.

The relationship between portal venous pressure and the degree of portasystemic shunting was studied in portal vein-ligated and cirrhotic rats anesthetized with halothane. One day after partial portal vein ligation there was a strong positive correlation (r = 0.80, n = 7) between portal pressure and shunting of mesenteric venous blood as measured by injection of radioactive microspheres. The relationship subsequently underwent rapid change but stabilized by 14 days postligation, when higher levels of shunting were again associated with higher portal pressures up to a limit of approximately 70% shunting, above which pressures did not increase further. This relationship was well described by a quadratic function (r = 0.75, n = 17). In cirrhotic rats there was no relationship between portal pressure and shunting (r = -0.01, n = 10). The results suggest that in the prehepatic model there is little inherent variability in capacity to develop shunts, which open to a degree directly related to portal pressure, but that this relationship may be altered in cirrhotic portal hypertension.

Source of raised serum estrogens in male rats with portal bypass.

We sought to establish the mechanism for the raised serum estrogen levels that occur in male rats with portal hypertension and resultant portal bypass. Using the portal vein ligated (PVL) rat model, we evaluated plasma steroid hormone concentrations, metabolic clearance rate (MCR) of estradiol, and hepatic metabolism of androstenedione to estrogens and other products. In contrast to serum testosterone levels that were reduced, serum androstenedione levels were normal in the PVL rat. Estradiol MCR was measured by a constant intravenous infusion technique and was found to be similar in PVL and control animals. Androstenedione MCR was determined during constant intravenous infusion of [3H]androstenedione, and the resultant radiolabeled steroids present in plasma were separated by thin layer chromatography. The MCR of androstenedione was not diminished in PVL rats compared with controls. However, there was a sevenfold increase in the plasma estradiol derived from [3H]androstenedione in rats with portal bypass. Examination of radiolabel excreted in bile during infusion of [3H]androstenedione showed that 25-46% of this steroid was converted to estradiol in PVL rats compared with less than 3% in control male rats (P less than 0.001). Moreover, there was a selective reduction in the excretion of 16 alpha-hydroxyandrostenedione, a finding which suggested that the metabolism of androstenedione via this pathway was decreased. Androstenedione 16 alpha-hydroxylation is known to be catalyzed by a male-specific cytochrome P-450 isoform, P-450UT-A. We conclude that raised plasma estradiol levels after portal bypass in male rats are due to increased production rates, resulting in turn from enhanced aromatization of androstenedione to estradiol. On the basis of the observed specific changes in androstenedione hydroxylation pathways, it is proposed that alterations in levels of sex-specific forms of cytochrome P-450 occur in male rats with portal bypass and could account for the enhanced formation of estradiol.

Glucagon and insulin metabolism in a portal-hypertensive rat model.

The role that portosystemic shunting plays in inducing the alterations of glucagon and insulin metabolism, which are observed in chronic liver disease, was studied in a rat model of prehepatic portal hypertension induced by portal vein constriction. Net splanchnic output of the hormones into the portal circulation was calculated from the difference between portal and systemic concentrations multiplied by portal plasma flow. Metabolic clearance rate was calculated as the ratio between output and systemic concentration. Portal blood flow was measured by the radioactive microsphere technique. Glucagon output in the portal vein-ligated rats was higher than in the sham-operated controls (5.9 +/- 1.5 vs. 2.0 +/- 0.2 ng/min, P less than 0.05). The metabolic clearance rate of glucagon was not significantly different between the two groups. Insulin output was not significantly different between the two groups; however, the metabolic clearance rate of insulin in the portal vein-ligated rats was reduced in comparison with the sham-operated group (9.5 +/- 1.5 vs. 18.4 +/- 3.3 ml/min, P less than 0.05). Our results indicate that portosystemic shunting per se is sufficient to cause an increased splanchnic output of glucagon into the portal system and a decreased metabolic clearance of insulin.

Splanchnic hemodynamics in portal-hypertensive rats: measurement with gamma-labeled microspheres.

A method to quantitate hepatic arterial flow (HA), portal venous flow (PBF), and blood flow through portal-systemic shunts (ShBF) in portal-hypertensive rats is described. This method relies on the injection of two differently radiolabeled microspheres (15 micrometers) into the left ventricle and spleen. To evaluate the usefulness of this technique, studies were performed on normal, cirrhotic, and portal vein-ligated rats anesthetized with ketamine. With this method, PBF is calculated indirectly from the sums of the blood flow of the splanchnic organs that drain into the portal vein. In the portal-hypertensive animals with portal-systemic shunting, this technique allows for the determination of PBF perfusing the liver [hepatic fraction of portal flow (HFP)] and PBF escaping through portal-systemic shunts (ShBF). The portal vein-ligated rats have higher HA flow (0.68 +/- 0.08 ml . min-1 . g-1) and lower HFP (0.08 +/- 0.01 ml . min-1 . g-1) than either the cirrhotic (HA: 0.27 +/- 0.03 ml . min-1 . g-1, P less than 0.01; HFP: 1.20 +/- 0.20 ml . min-1 . g-1, P less than 0.01) or the normal rat (HA: 0.29 +/- 0.06 ml . min-1 . g-1, P less than 0.01; HFP: 1.39 +/- 0.16 ml . min-1 . g-1, P less than 0.01). No significant difference was found between the cirrhotic and normal rats. The ShBF was higher in the portal vein-ligated rats (21.4 +/- 2.8 ml/min) than in the cirrhotic (4.6 +/- 2.5 ml/min, P less than 0.001) or normal rats (0.03 +/- 0.005 ml/min, P less than 0.01). The difference between the cirrhotic and normal animals was also significant (P less than 0.05). This is a simple, rapid, and reliable technique that allows for the quantitation of splanchnic hemodynamics in experimental models with portal hypertension.