Daniel F. Schafer

Reactivation of Chronic Hepatitis B Virus Infection by Cancer Chemotherapy

Two patients referred for cancer chemotherapy were found to be chronic, asymptomatic hepatitis B surface antigen (HBsAg) carriers. They had normal serum aminotransferase levels, but their sera were positive for HGsAg and antibody to hepatitis B e antigen. Both patients developed acute, icteric hepatitis within 3 months of starting cycled chemotherapy. In both cases, the disease seemed to be caused by a recurrence of type B hepatitis; it was accompanied by a marked increase in HBsAg titer and the appearance of hepatitis B virus DNA and DNA polymerase in the serum. One patient had a second episode of acute hepatitis after a second course of chemotherapy, but both patients ultimately recovered and became seronegative for HBsAg. Thus, it seems that cancer chemotherapeutic agents can reactivate type B hepatitis in asymptomatic HBsAg carriers. This reactivation is most likely due to an increase in hepatitis B virus synthesis followed by a rebound in host immune responses to hepatitis B virus infection when therapy is stopped. Such a phenomenon could have important implications for the therapy of chronic hepatitis B virus infection.

Randomized trial of chlorambucil for primary biliary cirrhosis

Twenty-four patients with primary biliary cirrhosis were entered into a prospective, randomized trial of chlorambucil therapy. Thirteen patients received chlorambucil (0.5-4 mg/day) and 11 patients received no therapy; all have been followed for 2-6 yr (mean, 4.1 yr). Two control but no treated patients died. Average serum bilirubin, serum aspartate aminotransferase activities, and albumin levels improved or remained unchanged in treated patients but worsened in controls. Serum alkaline phosphatase levels did not change in either group. Immunoglobulin M levels decreased and became normal in all treated patients but in only 3 control patients. Liver biopsy histology revealed an improvement in inflammatory cell infiltrate in treated patients in comparison with controls, but no significant change in degree of fibrosis or the histologic stage of disease. Side effects of therapy included bone marrow suppression necessitating discontinuation of the drug in 4 patients. These findings indicate that chlorambucil therapy may retard the progression of primary biliary cirrhosis. Whether such therapy will ultimately decrease morbidity and improve survival in this disease can only be demonstrated by large-scale, placebo-controlled trials.

Visual Evoked Potentials in a Rabbit Model of Hepatic Encephalopathy: I. Sequential Changes and Comparisons With Drug-Induced Comas

The recording of visual evoked potentials in rabbits has been shown to be an objective, reproducible, noninvasive technique for quantitating changes in the pattern of cerebral neuronal activity. The development of hepatic encephalopathy due to galactosamine-induced fulminant hepatic failure was consistently associated with a series of distinctive changes in the visual evoked potential waveform. The pattern of the visual evoked potential in hepatic coma (due to galactosamine-induced fulminant hepatic failure) differed fundamentally from that in ether-induced coma, but was identical to that in comas induced by three drugs which activate gamma-aminobutyric acid-ergic neural mechanisms: pentobarbital, diazepam, and muscimol. These findings are compatible with activation of the gamma-aminobutyric acid inhibitory neurotransmitter system contributing to cerebral neuronal inhibition in hepatic coma due to galactosamine-induced fulminant hepatic failure.

Visual evoked potentials in a rabbit model of hepatic encephalopathy. II. Comparison of hyperammonemic encephalopathy, postictal coma, and coma induced by synergistic neurotoxins

To assess neuronal mechanisms of potential importance in the pathogenesis of hepatic encephalopathy, visual evoked potentials were recorded in rabbits with acute hyperammonemic encephalopathy, postictal coma, and toxin-induced coma resulting from the administration of a combination of subcoma doses of three neurotoxins: ammonia, dimethyldisulfide, and octanoic acid. The patterns of visual evoked potentials in these three syndromes were compared with those of rabbits with hepatic encephalopathy due to galactosamine-induced fulminant hepatic failure. In the absence of seizures, the patterns of visual evoked potentials associated with hyperammonemic encephalopathy and toxin-induced coma were fundamentally different from those associated with any stage of hepatic encephalopathy due to galactosamine-induced fulminant hepatic failure. In contrast, the pattern of visual evoked potentials in early postictal coma induced by four different precipitating factors (including toxin-induced seizures) resembled that of late-stage hepatic encephalopathy due to galactosamine-induced fulminant hepatic failure. These findings suggest that the recording of visual evoked potentials may be of value in experimentally testing hypotheses of the pathogenesis of hepatic encephalopathy due to fulminant hepatic failure. They indicate that acute hyperammonemia is not a satisfactory model of hepatic encephalopathy due to galactosamine-induced fulminant hepatic failure, that the occurrence of seizures may lead to incorrect interpretation of experimental data from models of hepatic encephalopathy, and that the syndromes of hepatic encephalopathy due to galactosamine-induced fulminant hepatic failure and postictal coma may share similar neural mechanisms. Finally, the results of this study do not support the hypothesis that hepatic encephalopathy due to galactosamine-induced fulminant hepatic failure is mediated by the synergistic interaction of ammonia, mercaptans, and fatty acids on the brain.

Colonic bacteria: A source of γ-aminobutyric acid in blood

Abstract Blood contains more of the major inhibitory neurotransmitter, γ-aminobutyric acid (GABA), than does cerebrospinal fluid. However, the source and physiologic importance of this compound in the peripheral circulation are unknown. We have found that blood drawn from the portal vein of normal rabbits contains twice the concentration of GABA as that drawn simultaneously from the abdominal aorta. Also, human-derived colonic bacteria grown anaerobically secreted a large amount of GABA equivalents into culture media when measured by a radioreceptor technique. The production of inhibitory neuroactive substances, such as GABA, by colonic bacteria may play a role in the pathogenesis of certain disease states, for example, hepatic encephalopathy.

Selective immimoglobulin a deficiency associated with primary biliary cirrhosis in a family with liver disease

A family is described in which multiple members are afflicted with liver disease and primary biliary cirrhosis (PBC). In the third generation, one member died of PBC, and a second individual has both symptomatic PBC and selective immunoglobulin A (IgA) deficiency, an association not previously reported. By culturing this patients lymphocytes in vitro it was shown that the IgA deficiency was due to a failure of B cells to secrete IgA. Two other siblings of this patient have multiple serum biochemical and serologic abnormalities that are sometimes associated with PBC, but they do not have histopathologically overt PBC or IgA deficiency. All three surviving family members have a diminished autologous mixed lymphocyte reaction, an immunologic abnormality that has previously been found in patients with PBC, selective IgA deficiency, and several autoimmune diseases. As there is an association between selective IgA deficiency and certain autoimmune diseases, it is possible that this immunodeficiency contributed to the development of PBC in the patient in whom the two diseases coexisted. Furthermore, the occurrence of PBC in a patient with selective IgA deficiency indicates that the pathogenesis of PBC does not require IgA-dependent immune mechanisms.

Enzymes of Cerebral GABA Metabolism and Synaptosomal GABA Uptake in Acute Liver Failure in the Rabbit: Evidence for Decreased Cerebral GABA‐Transaminase Activity

Abstract: Measurements of the activities of the two key enzymes in cerebral GABA metabolism—glutamate decarboxylase (GAD) and GABA‐transaminase (GABA‐T)—were performed in normal rabbits and in rabbits with hepatic encephalopathy due to galactosamine‐induced liver failure. Furthermore the uptake of GABA by synaptosomes was studied. Hepatic encephalopathy was associated with a marked decrease in the activity of GAB A‐T. This decrease in activity was already apparent in galactosamine‐treated rabbits before the onset of hepatic encephalopathy. Sera and serum ultrafiltrates of rabbits with hepatic encephalopathy but not of normal rabbits or of rabbits with uremic encephalopathy were shown to inhibit GABA‐T activity in vitro. Cerebral GAD activity and synaptosomal GABA uptake in rabbits with hepatic encephalopathy and in untreated animals were not different. These later findings indicate that hepatic encephalopathy is not associated with alterations of presynaptic GABA nerve terminals in the central nervous system. The demonstration of a decrease in cortical GABA‐T activity provides indirect evidence for decreased GABA turnover in the brains of rabbits with hepatic encephalopathy and thus is compatible with augmented GABA‐ergic inhibitory neurotransmission contributing to the neural inhibition of hepatic encephalopathy.

1581 Blood-brain barrier changes in acute hepatic encepha-lopathy

Blood-to-brain transport constants (k) for 14C-α-aminoisobutyric acid (AIB), a marker of brain capillary permeability, were measured in rabbits with acute hepatic encephalopathy induced by the hepatotoxin, galactosamine (GAL). Tissue 14C-activity was assayed on 20 micron thick coronal sections of frozen rabbit brains by quantitative autoradiography; adjacent sections were prepared for histological examination. Experiments which were performed 2 and 10 hours after the administration of GLA (times when the rabbits were neurologically normal) showed no changes in the k values of AIB from the untreated controls. Eighteen hours after GAL administration, the rabbits were comatose and their brains were edematous; in addition, five-to tenfold increases in blood-brain barrier (BBB) permeability to AIB were found in gray matter (no permeability changes were detected in white matter). These observations suggest that the permeability of the BBB not only to AIB but also to other solutes (notably, Na and Cl) and to water has increased in gray matter and that the role of the BBB as a regulatory of brain volume has been compromised by these changes, thereby leading to cerebral edema. A similar pathogenesis may be involved in other diseases in which hepatic dysfunction frequently leads to acute cerebral edema, elevated intracranial pressure, and death (for example, Reyes syndrome).

Hepatic Encephalopathy and Benzodiazepine Receptor Ligands

Benzodiazepine (BZ) receptor ligands interact with specific binding sites on the GABAA receptor/chloride channel complex on neurons in the central nervous system (CNS) (1) (Figure 1). Such ligands could contribute to or ameliorate hepatic encephalopathy (HE) if the supra-molecular complex itself is involved in this syndrome. The complex includes distinct receptors for BZ ligands and GABA, and a chloride ionophore on which there are binding sites for barbiturates and cage convulsants, such as Picrotoxin. Activation of the effector component of this complex, the chloride channel, is mediated by the binding of GABA to its receptor and is potentiated by the binding of a BZ agonist or a barbiturate to discrete loci on the complex. Activation induces conformational changes in the complex which result in opening of the chloride channel and hyperpolarization of the neuron (1) (Figure 1). This GABA-gated chloride ion conductance is the basis of GABA-mediated inhibitory neurotransmission or “GABAergic tone.”

A nonlinear kinetic model of γ-aminobutyric acid metabolism in a rabbit model of acute liver failure

Abstract To investigate the mechanisms by which serum levels of γ-aminobutyric acid (GABA) become elevated in experimental acute liver failure, a multicompartmental model of GABA metabolism has been constructed and used to simulate previously generated data on the kinetics of 3H-GABA uptake by isolated hepatocytes from normal rats and the kinetics of 3H-GABA in the plasma of normal rabbits, rabbits with galactosamine-induced acute liver failure, and rabbits with divascularized livers. Modeling analysis revealed that acute liver failure was associated with values for the mean fractional catabolic rate of GABA, plasma volume, and hepatic extraction of GABA that were 29%, 12%, and 49% less, respectively, than the corresponding control values. The defect in hepatic tissue extraction of GABA was sufficient to account for only 60% of the 10-fold increase in serum GABA levels that occurs in acute liver failure. Furthermore the 10-fold increase in serum GABA levels occured in acute liver failure before the onset of overt hepatic encephalopathy when hepatic extraction of GABA was not appreciably different from that found in normal rabbits. Thus the increase in serum GABA levels that occurs in acute liver failure cannot be attributed to a defect in hepatic extraction of GABA alone. Indeed, the modeling analysis indicated that in acute liver failure there is a 3—8-fold increase in the rate of delivery of GABA to the systemic circulation, but did not indicate its source.