cally prevents liver injury from this or other causes. This No published reports compare hepatic a-tocopherol (adcontroversy is the result of many factors, including use of justed for hepatic lipid content) with indicators of blood different dosages of vitamin E, differences in the mechanisms a-tocopherol in adult patients with various liver diseases.of liver injury between experimental animals and clinical a-Tocopherol was simultaneously measured in liver biopsy subjects, and the lack of a reliable indicator of hepatic vitamin tissues and blood from 66 subjects (9 comparison patients E status. Because vitamin E-deficient red blood cells (RBCs) hospitalized for biliary tract surgery, 13 with chronic persisare fragile (as a result of hydrogen peroxidation), 7 RBCs are tent hepatitis, 9 with chronic aggressive hepatitis, 10 with considered a peripheral target tissue for vitamin E. RBC aacute hepatitis, 10 with cirrhosis, 7 with both cirrhosis and tocopherol has been assumed to be a reliable indicator of hepatic cell carcinoma, and 8 with fatty liver). Hepatic, erythvitamin E status in biomembranes. 8,9 The ratio of plasma arocyte, and plasma a-tocopherol concentrations were meatocopherol to total serum lipid concentration (plasma Toc/ sured, as were hepatic and serum lipids. The ratios of a-TL ratio) has also been used as a clinical index of plasma tocopherol to total lipid concentrations (Toc/TL ratios) in vitamin E status. 7,10-12 However, to our knowledge, no pubplasma and liver were calculated. In both comparison patients lished reports compare hepatic a-tocopherol concentrations and patients with chronic persistent hepatitis and fatty liver, with RBC a-tocopherol concentrations or plasma Toc/TL rahepatic a-tocopherol concentrations were strongly correlated tios in adult patients with liver disease. with hepatic triglyceride and total lipid concentrations (r Å Previous reports have shown diminished hepatic a-to-.72, P õ .001; and r Å .75, P õ .001, respectively); the copherol concentrations in alcoholic patients with cirrhorelationships (slopes) when hepatic a-tocopherol concentrasis, 13,14 but another report showed no correlation between tions were compared with hepatic triglyceride and total lipid the ratios of hepatic a-tocopherol to total hepatic lipid conconcentrations were similar in these patients and in all subcentrations (hepatic Toc/TL ratios) and inflammatory histojects. No strong correlations were found between hepatic and logical changes in alcoholic patients. 15 In our study (Nagita blood a-tocopherol parameters in all subjects. These results A, et al., unpublished observations, January 1997), hepatic suggest that hepatic a-tocopherol is present at similar concena-tocopherol concentrations were higher in pediatric patrations in triglycerides as well as total cholesterol and phostients with fatty liver (FL) than in comparison subjects and pholipids and that neither plasma Toc/TL ratios nor erythrowere strongly correlated with hepatic areas occupied by fat cyte a-tocopherol concentrations are usefu...