Pathogenesis of precirrhotic portal hypertension in alcohol-fed baboons.

To study mechanisms and anatomic correlates of precirrhotic portal hypertension, we measured portal pressure either at laparotomy (in the portal vein) or by hepatic vein catheterization (wedge pressure) in 24 pairs of baboons fed 50% of energy either as ethanol or isocaloric carbohydrate (controls) for 4 mo-9 yr. On liver biopsy 7 had simple fatty liver; none had portal pressure exceeding the control range (2.7-13.0 cmH2O). The remaining 17 alcohol-fed baboons had fibrous tissue deposition around the terminal hepatic venules and adjacent sinusoids. The mean portal pressure was significantly increased (15.0 +/- 1.4 cmH2O) compared with the value in baboons with fatty liver (9.6 +/- 0.9 cmH2O) and in controls (8.0 +/- 0.6 cmH2O), with 8 animals exceeding the control range. Estimated hepatic blood flow was unchanged. Alcohol feeding resulted in increased hepatocyte size in both the fatty liver and fatty liver with fibrosis group; however, portal pressure did not correlate with alterations of cell size, liver volume, hepatic triacylglycerol, and protein contents. By contrast, for veins of comparable size, there was a significant correlation (r = 0.6666, p less than 0.01) between the thickness of the perivenular fibrous rim and portal pressure. Perivenular fibrosis was commonly associated with adjacent perisinusoidal fibrosis and this lesion also correlated with portal pressure. Furthermore, if one postulates that increased cell size causes enhanced pressure with secondary fibrosis, the latter should have first occurred "upstream," in the mid and portal zones. Sequential biopsy specimens, however, showed that fibrosis first appeared in the perivenular areas, suggesting that, in most instances, increased pressure is in fact secondary to the perivenular fibrosis.