Hepatic sinusoidal fibrosis induced by cholesterol and stilbestrol in the rabbit: 1. Morphology and inhibition of fibrogenesis by dipyridamole.

This study documents the hepatic morphology and the ultrastructure of a model of hepatic fibrosis in rabbits. Rabbits were given a cholesterol-supplemented diet (1%), a stilbestrol diet (10 mg subcutaneously twice a week), or both treatments simultaneously for 7 weeks. Rabbits given the combined treatment developed sinusoidal and portal fibrosis with only a mild disturbance of acinar vascular relationships. Ultrastructurally, there was marked widening of the spaces of Disse by collagen fibers, basement membrane material adjacent to endothelial cells and hepatocytes, blunted hepatocellular microvilli, activated stellate cells, lipid droplets in endothelial cells and hepatocytes, and degranulated platelets in sinusoids. The hepatic hydroxyproline content was markedly increased (12.0 +/- 5.2 vs. 4.8 +/- 1.5 mmol/g of liver dry weight; P < .001). Plasma bile acids were markedly increased (222 +/- 180 vs. 12 +/- 5 in controls; P < .001). Dipyridamole (25 mg every 12 hours) that was given in addition to cholesterol and stilbestrol decreased the hepatic collagen content (-49% and -48%, in two experiments; P < .05 in both) and splenomegaly. This model provides a reliable method for the production of extensive sinusoidal fibrosis with capillarization of sinusoids. Hepatocellular degeneration is only mild to moderate, and fibrosis occurs slowly without the sudden pathological changes that occur with other models of hepatic fibrosis, such as with the administration of CCl4 or galactosamine. The mechanism of injury may involve the accumulation of bile salts or the generation of free radicals from cholesterol oxidation products. The possibility that the sinusoidal release of platelet-derived factors may have a role in the activation of stellate cells (lipocytes) is supported by the suppression of fibrogenesis by dipyridamole.

The effect of alcohol-induced hepatomegaly on portal hypertension in cirrhotic rats.

Male Sprague-Dawley rats with CCl4-induced cirrhosis (confirmed by increased collagen content and light microscopy) were fed either ethanol (Group A, n = 9) or isocaloric carbohydrate diet (Group B, n = 8) for 4 weeks. Histologic and hemodynamic measurements were obtained in the awake state before (time 1) and after the 4 weeks of diet (time 2). Portal-systemic shunts were evaluated using radiolabelled microspheres. Liver weight was increased in Group A (16.5 +/- 0.5 vs. 14.2 +/- 0.5 g, mean +/- SE, p less than 0.005) as was the ratio of liver weight over total body weight (3.41 +/- 0.05 vs. 2.86 +/- 0.09%, p less than 0.0001, +19.2%). Hepatocytes surface area was increased in the ethanol group (357 +/- 9 vs. 294 +/- 7 microns 2, p less than 0.0001). In Group B, only 9 +/- 2% of hepatocytes had steatosis as opposed to 69 +/- 3% of centronodular and 34 +/- 3% of perinodular hepatocytes in Group A (p less than 0.001). Portal pressure remained stable in both groups (time 1 (A) 16.9 +/- 0.8, (B) 15.8 +/- 1.1 mmHg, n.s.; time 2 (A) 15.9 +/- 0.7, (B) 15.8 +/- 0.6 mmHg, n.s.). Portal-systemic shunts did not change with time or diet (time 1 (A) 10.6 +/- 3.7%, (B) 4.1 +/- 2.1%, n.s.; time 2 (A) 13.4 +/- 5.9%, (B) 10.8 +/- 4.3%, n.s.).(ABSTRACT TRUNCATED AT 250 WORDS)

The rat liver microcirculation in alcohol-induced hepatomegaly.

It has been suggested that hepatocyte enlargement can lead to compression of the extracellular space (sinusoidal and interstitial) and induce portal hypertension. However, this hypothesis has never been tested by measuring the vascular and extravascular spaces in the intact liver. The aim of the present study was to investigate the effects of chronic alcohol intake on the hepatic microcirculation using Goresky's multiple-indicator dilution technique in the isolated perfused rat liver. Female rat littermates were pair-fed either ethanol (n = 7) or an isocaloric carbohydrate diet (n = 7) for 21 days. As expected, chronic alcohol intake produced a significant increase in liver/body weight ratio (+32%, p less than 0.01) and hepatocyte size (+45%, p less than 0.001), which was accompanied by a marked increase in the cellular water space (control: 3.3 +/- 0.6 ml; ethanol-fed: 4.9 +/- 0.9 ml; p less than 0.001). When expressing data per total liver, the sinusoidal space was similar in the two groups (control: 1.87 +/- 0.2; ethanol-fed: 1.95 +/- 0.2 ml; not significant), whereas the interstitial space was increased in alcohol rats compared to controls (albumin space +58%, p less than 0.01; sucrose space +51%, p less than 0.01). In alcoholic rats, the sinusoidal space was probably stretched, with an overall reduced transversal diameter, as suggested by the reduced values found when data were expressed per gm of liver weight. However, despite this finding and the enlargement of the liver and hepatocytes observed in alcoholic rats, similar values were obtained between the two groups for the portal perfusion pressure and thus the intrahepatic vascular resistance.(ABSTRACT TRUNCATED AT 250 WORDS)

Estimation of collagen content of liver specimens. Variation among animals and among hepatic lobes in cirrhotic rats.

We undertook a study to evaluate the correlation between morphometric evaluation and colorimetric determination of hepatic collagen content, and to analyze the variation among animals as well as among lobes of the same liver in hepatic collagen content after CCl4-induced micronodular cirrhosis. The results revealed a significant correlation (r = 0.9458; p less than 0.001) between the morphometric and colorimetric methods of collagen evaluation of liver specimens; both methods also significantly distinguished data obtained from controls and from cirrhotic rats (p less than 0.0005). After induction of micronodular cirrhosis by chronic CCl4 administration, a highly significant variation in hepatic collagen content was observed among animals (p less than 0.0001). By contrast, no significant difference in collagen content was observed (p less than 0.05) among hepatic lobes of a given animal. These results indicate that in this animal model of liver cirrhosis, interpretation of biochemical data would benefit by being related to the severity of the hepatic collagen infiltration of each animal. Our data also show that representative values for total hepatic collagen infiltration can be obtained from a single liver specimen; we suggest, however, that the specimen be taken from a major lobe of the liver and that a sufficiently large number of animals be used to avoid occasional sampling errors.