The aim of this study was to evaluate whether there plain, for instance, the so-called hyperdynamic circulais endothelial dysfunction in patients with cirrhosis and tion, which is characterized by hypotension, low vascuto detect the mechanism that may account for it. We lar resistance, and increased cardiac output and is measured plasma levels of von Willebrand factor (vWF), thought to be attributable to increased vascular release a marker of endothelial perturbation, and endotoxin, of nitric oxide, 3,4 a vasodilating substance secreted by which releases vWF from endothelial cells in vitro, in endothelial and vascular smooth muscle cells.5 Endo-32 patients (18 men, 14 women, aged 39-70 years) with thelial perturbation might also explain the coagulation cirrhosis classified as mild (class A, n Å 10), moderate changes seen in cirrhosis. It is known that the clinical In cirrhosis, changes in hemostasis might be medibetween vWF antigen and endotoxemia (r Å .92; P ated by endotoxemia, which is elevated perhaps as a Å .0001). In 20 selected patients, vWF antigen and endoresult of reduced hepatic clearance.12 There is evidence toxemia were measured before and after 7 days of stanthat endotoxemia induces vascular secretion of nitric dard therapy (n Å 10) or standard therapy plus nonabsorbable antibiotics. There was a significant decrease of oxide and enhances activation of clotting. An in vitro vWF antigen (P õ .02) concomitantly with the decrease study demonstrated that endotoxin enhances nitric oxof endotoxemia (P õ .006) in patients taking nonabsorb-ide synthase expression in endothelial cell, 13 whereas able antibiotics. Human umbilical vein endothelial cells an in vivo study showed that in cirrhotic patients the incubated in vitro with 125 to 500 pg/mL endotoxin re-reduction of endotoxemia by nonabsorbable antibiotics leased vWF antigen into the medium dose dependently. is associated with a significant decrease of serum levels These results demonstrate that there is endothelial per-of nitrite and nitrate.2 In addition, in experimental turbation in cirrhosis and that endotoxemia may play a studies in which endotoxin was infused into humans, key role in its occurrence. (HEPATOLOGY 1996;23:1377-there was a significant increase of the prothrombin 1383.) fragment 1 / 2, 14 a marker of in vivo thrombin generation, suggesting that endotoxin may heighten coagulaThere is indirect evidence that there may be endothetion enzyme activity. lial perturbation in cirrhosis, which might explain All of these data indicate that endotoxemia is the some clinical and biological changes occurring in this trigger of a sequence of events inducing endothelial, clinical setting. karyocytes, which is considered to be a marker of endoAddress reprint requests to: Francesco Violi, M.D., I Clinica Medica, Polithelial perturbation. 15 We also attempted to reduce the