Search

Clinically evident, generalized graft-vs.-host disease is not thought to occur as a consequence of transplantation of most solid organs. The large inoculum of donor lymphoid cells required and the inability of the recipient to destroy these passenger cells are apparently rarely produced by solid-organ transplantation. However, there is a risk of graft-vs.-host disease when normal spleens are included in vascularized pancreas transplants. We report on a case of severe, generalized acute graft-vs.-host disease in a recipient of a liver tranplant, involving documented transient chimerism of donor origin. This life-threatening complication was treated successfully with increase in the corticosteroid dosage and administration of anti-thymocyte globulin.

Editor,—Savasan et al describe the diverse immune abnormalities, limited effect of different therapeutic manoeuvres (splenectomy, steroids, intravenous immunoglobulins), and poor prognosis in patients with autoimmune thrombocytopenia and Coombs positive haemolytic anaemia (Evans’ syndrome).1 Fifty five per cent of their patients (six out of 11) had hepatomegaly, some with portal triaditis. We would like to extend their findings by reporting our recent experience of a patient with Evans’ syndrome who also had Coombs positive giant cell hepatitis (CPGCH). CPGCH is a rare disease of early childhood, with unknown pathogenesis and variable response to immunosuppression.2Liver transplantation is not curative, since the disease recurs in the graft.3 An Asian girl, born from consanguineous parents, was diagnosed …

Lymphocytes from 39 patients with HBsAg positive chronic liver disease were incubated with their own hepatocytes to investigate mechanisms of lymphocyte-mediated liver damage. Cytotoxicity was significantly increased in 46% overall, and in 73% of those with chronic active hepatitis. Unlike HBsAg negative chronic active hepatitis where only non-T cells were cytotoxic, HBsAg positive patients had both cytotoxic T and non-T cells. A purified liver membrane complex (LSP) and aggregated IgG both blocked non-T cytotoxicity without affecting T cell cytotoxicity; this suggests that the former is probably an antibody-dependent cell-mediated reaction against normal membrane components. This was confirmed in preliminary studies which demonstrated that preincubation of hepatocytes with the F(ab)2' fragment of an anti-human IgG reduced non-T lymphocyte cytotoxicity. T-cell cytotoxicity was restricted to HBeAg-positive patients, suggesting a relationship between T-cell cytotoxicity and viral replication. Purified HBsAg, however, blocked cytotoxicity in only three of 11 cases. Non-T lymphocytes reacting with normal membrane components may contribute to liver damage in both 'autoimmune' and virus-associated chronic liver disease, whereas cytotoxic T-cells, probably reacting with viral determinants, are exclusive to those with viral replication.

Virus-like particles (60-70 nm) with spiked surfaces budding into cell vacuoles and rod-shaped inclusions were detected in nuclei of hepatocytes from a British patient transplanted for sporadic non-A, non-B fulminant hepatitis (NANB-FHF), probably contracted in Kenya. Identical particles were seen in two successive grafts (days 2 and 10) at regrafting for recurrent FHF. Ultrastructural features resembled those of the RNA-containing arbovirus, Rift Valley fever virus, but serological markers against a representative panel for arboviruses (Togaviruses) and transmission in mice proved negative. The particles shared features with the different arboviruses seen in the hepatectomy specimen of a second patient with NANB-FHF, and in both patients an insect vector was implicated in the clinical history. The particles were identical in size to those of a third patient with NANB-FHF, who had remained in the United Kingdom. These findings, together with the recent report of isolation of an RNA-containing virus resembling the Togaviridae, in parenteral NANB, suggest that several exotic virus-like agents resembling the arboviruses may be involved in the aetiology of NANB, including in the sporadic forms of FHF in the United Kingdom.

Of 98 patients dying with primary biliary cirrhosis only four developed hepatocellular carcinoma. It is suggested that the development of hepatocellular carcinoma is uncommon in this type of chronic liver disease because of its known female preponderance, and the fact that cirrhosis develops late in the course of the illness.

Double-shelled virus-like particles (60 nm) and long cytoplasmic tubular structures were found in the cytoplasm of hepatocytes from areas of collapsed and regenerating areas of hepatectomised liver in a 13-year-old boy who received a liver graft for fulminant hepatitis attributed to sporadic non-A, non-B hepatitis. The patient died on the ninth postoperative day from acute graft failure. Although virus-like particles were not found, instead, gram-negative rods were identified in the necrotic graft and the most likely cause of death was a gram-negative septicaemia with a Shwartzman-like reaction localized to the liver.

The effect of azathioprine on survival of patients with primary biliary cirrhosis was studied prospectively in a multinational, double-blind, randomized clinical trial including 248 patients of whom 127 received azathioprine and 121 placebo. There were 57 deaths in the azathioprine group and 62 in the placebo group. The actual survival was slightly longer during azathioprine than during placebo treatment. Using Cox multiple regression analysis and adjusting for slight imbalance between the two treatment groups, the therapeutic effect of azathioprine was statistically significant (p = 0.01), with azathioprine reducing the risk of dying to 59% of that observed during placebo treatment (95% confidence interval 40%-90%) or improving survival time by 20 mo in the average patient. Furthermore, azathioprine slowed down progressing incapacitation. Side effects of azathioprine were relatively few. The analysis revealed that the following five variables independently implied poor prognosis: high serum bilirubin, old age, cirrhosis, low serum albumin, and central cholestasis. These factors were combined to a "prognostic index" for prediction of outcome in new patients. The index was validated on independent patient data. On the basis of these results we recommend azathioprine as a routine treatment of primary biliary cirrhosis.

One hundred and eighty nine patients with primary biliary cirrhosis were entered into a double blind, placebo controlled randomised trial starting in January 1978 to assess the therapeutic value of d-penicillamine 1200 mg daily. Eighteen of the 98 patients receiving d-penicillamine and 22 of the 91 placebo treated patients died during the study. Thirty six per cent of those on d-penicillamine and 8% of those on placebo were withdrawn from the study. No difference in overall survival was noted between the two groups of patients whether the results were analysed for the entire period of observation or only during the period in which the patients were receiving therapy. The mortality rate of those receiving d-penicillamine in histological stage I to II, however, was one third of that of the placebo group although this difference did not reach statistical significance. Using the occurrence rate ratio as the statistical method of analysis, no effect of d-penicillamine was noted on any clinical, biochemical or histological features examined, except the serum alanine aminotransferase activity which was greater in those on active treatment. In this trial we have been unable to establish any therapeutic benefit from the drug.

To assess whether hepatic peptidyl prolyl hydroxylase (PPH) activity could serve as a practical quantitative indicator of hepatic fibrosis or aid in the categorization, diagnosis or prognosis of hepatobiliary disorders in infancy and childhood, the activity of this enzyme has been determined prospectively by a tritium release method in 97 biopsies from 94 infants and children with the following conditions: acute hepatitis of infancy, 10 patients; extrahepatic biliary atresia, 13; previous hepatitis of infancy, 8; α-1-antitrypsin deficiency, 6; chronic active hepatitis, 17; chronic persistent hepatitis, 5; glycogen storage disease, 5; and 25 patients with a miscellanea of other liver disorders. PPH activity was considered in relation to diagnosis, biochemical and histological abnormality and subsequent prognosis over a 4-year period. Five liver biopsies which showed no histological abnormality were considered as “controls” having PPH values of 0.72 ± 0.47 (mean ± S.D.). PPH activity was significantly elevated in acute hepatitis of infancy, 9 of the 10 infants having PPH >1.66 units (i.e., mean ± 2 S.D. of the “control” value). Nine infants (70%) with extrahepatic biliary atresia also had PPH activity above this value, as did two with a-1-antitrypsin deficiency and 12 patients all in different diagnostic categories. PPH activity did not correlate with hepatic fibrosis as indicated by hepatic hydroxyproline concentration or by histological assessment, or with biochemical tests of liver function within any diagnostic group or in the series as a whole. PPH activity was similar in biopsies with and without histological features of cirrhosis. Of the 32 patients with significantly elevated PPH activity, deterioration of their hepatic state in the subsequent 4 years occurred in only 10: 5 with extrahepatic biliary atresia and individual children with acute hepatitis of infancy, α-1-antitrypsin deficiency, biliary hypoplasia, cystic fibrosis and biliary cirrhosis associated with hypogammaglobulinemia. Thus, although elevated PPH activity may indicate an increase in hepatic collagen biosynthesis at the time of biopsy, it does not appear to be a prognostic indicator of progressive liver damage. We conclude that the determination of hepatic PPH activity is of no diagnostic or prognostic value in hepatobiliary disorders in childhood, and the need for a practical method of monitoring persisting or continuing accumulation of fibrous tissue within the liver remains.

The histological appearances of the liver damage occurring after a paracetamol overdose are described in liver biopsies from 104 patients, of whom 38 developed fulminant hepatic failure. Confluent centrilobular necrosis of varying extent was followed by rapid disappearance of necrotic cells, leaving areas of reticulin collapse and a usually mild inflammatory reaction. The histological recovery in even the most severe cases was remarkable, and only one of the 17 survivors whose initial biopsy showed the pattern of interlobular bridging necrosis had appreciable residual fibrosis in a follow-up biopsy taken after 1 yr. A quantitative estimate was made of the amount of surviving liver parenchyma using a morphometric technique and the hepatocyte volume fraction (HVF) in biopsies performed within 10 days of the overdose correlated well with the clinical course and both the maximal prolongation of the prothrombin time and the peak plasma bilirubin concentration in the first 10 days. An HVF value (normal 85 +/- 5 per cent.) of less than 40 per cent. in a biopsy taken within 10 days of the overdose was found only in patients who died. However, HVF measurements on biopsies from three survivors taken later than 10 days after the overdose shows that survival is possible below this critical level.

The relation between donor and recipient status for HLA class I and II antigens in 62 patients undergoing liver transplantation was examined with particular reference to a well-defined variant of chronic rejection, the vanishing-bileduct (VBD) syndrome. A complete mismatch for class I antigens was more common in those with the VBD syndrome than in those with normal graft function or chronic graft malfunction unrelated to the syndrome (p less than 0.025). In contrast, a complete mismatch for class II antigens was considerably less common in those with the VBD syndrome than in those without (p less than 0.02). The association of a complete mismatch for class I and a partial or complete match for class II antigens with the VBD syndrome was highly significant (p less than 0.0005). These findings support the hypothesis that in the VBD syndrome both class I antigen expression on bileduct epithelium and immunological interaction at the level of class II antigens are required for the rejection process to occur. In addition, high-titre donor-specific antibodies to class I antigen, which were present in 6 of 14 of those with the VBD syndrome but in none of those without (p less than 0.0005), may be involved in the pathogenesis of bileduct damage.

To determine the comparative value of serum Type III procollagen peptide (PIIIP) in paediatric and adult liver disease we have measured PIIIP in 201 healthy subjects (aged 1 day-77 years) and twenty-one children and five adults with chronic active hepatitis (CAH). Healthy children had significantly higher PIIIP levels than adults (P less than 0.001), with highest values of 298 +/- 88 ng ml-1 (s.d.) in the neonatal period. PIIIP fell to 30.9 +/- 7.0 by 1 year, 19.1 +/- 4.5 by 3 years and rose significantly (P less than 0.01) at puberty. Adult levels (8.3 +/- 3.2) occurred by 16 years of age. Serum PIIIP levels were significantly elevated (P less than 0.001) in adults when they had biochemical and histological evidence of active liver disease but were consistently within the normal range for age in 70% of children with similar hepatic pathology. The minor elevations in PIIIP in the other children were unrelated to clinical, biochemical or histological evidence of active liver disease. While raised PIIIP may be a non-invasive marker of liver disease activity in adults, its value in childrens' disorders appears to be limited by the high levels of PIIIP which occur during growth.

The contribution of cytomegalovirus (CMV) infection and its interrelation with HLA antigens in the development of chronic rejection (vanishing bile-duct syndrome--VBDS) was investigated in 101 patients surviving for at least 3 months after liver transplantation. A 1-2 antigen match for HLA DR antigens (30.9% vs 4.5% for zero DR match; p less than 0.002), a zero match for HLA A/B antigens (27.5% vs 10.9% for 1 or more A/B match; p less than 0.05), and active CMV infection (26.3% vs 4.4% for no CMV infection; p less than 0.005) were independently associated with an increased risk of VBDS. The coexistence of a 1-2 HLA DR match and CMV infection carried the highest relative risk (10.1) of VBDS; these two variables were probably interdependent since either alone was associated with a low relative risk (0.45 and 0.5). The association between VBDS and active CMV infection was not a consequence of alterations in immunosuppressive therapy. The findings would be consistent with precipitation of chronic rejection by CMV-induced HLA antigen expression in patients rendered susceptible by the donor/recipient HLA antigen match.

The leukocyte migration inhibition test has been used to investigate cellular immune responses to antigens in a protein fraction (BPF) of normal human gallbladder bile in patients with a variety of intraand extrahepatic diseases. Inhibition of leukocyte migration in the presence of BPF was observed in 30 (81%) of 37 patients with PBC, in 8 (80%) of 10 patients with sclerosing cholangitis, and in 7 (26%) of 27 patients with chronic active hepatitis. Only 1 of 31 patients with other liver diseases or with uncomplicated ulcerative colitis showed a similar response to BPF. The BPF was found to contain three antigens which were distinct from plasma proteins. Immuno-fluorescence studies revealed that one of these antigens appears to be derived from that part of the hepatocellular membrane which forms the bile canaliculus and that a second appears to be associated with the epithelial cell membranes of interlobular and septal bile ducts. The site of origin of the third antigen could not be established. It is suggested that cellular immune responses to biliary antigens could be involved in the progressive bile duct destruction of chronic biliary disease.

In a prospective study of 613 patients with cirrhosis of different aetiological types increasing age, male sex, and non-UK nationality were found to be significant independent risk factors for the progression of cirrhosis to hepatocellular carcinoma. Seropositivity for hepatitis B surface antigen and the duration and aetiology of cirrhosis were not related to the development of the disease. Age and sex were also found to be significant risk factors when UK patients alone were considered. Seropositivity for hepatitis B surface antigen may be associated with hepatocellular carcinoma only because hepatitis B virus infection is a common cause of cirrhosis.

Antibody to the hepatitis B core antigen (anti-HBc) was detected in sera from 6 (19%) of 32 patients with HBsAg-negative chronic active hepatitis. In three cases with the highest anti-HBc titers, core antigen was detected in liver cell nuclei and in one case HBsAg was also present in hepatocytes, suggesting continuing hepatitis B virus infection. During follow-up, anti-HBc titers fell slowly in those with no viral antigens in liver tissue, and in two cases with virus in the liver at presentation, viral antigens were no longer demonstrable four and eight years later. In one case, clearance of virus was preceded by the appearance of HBsAg in liver and serum, suggesting reactivation of viral replication. In three of the anti-HBc-negative cases a nuclear antigen unrelated to the hepatitis B virus was detected by immunofluorescence, and it is possible that liver disease in these patients may be related to persistent non-A, non-B virus infection.

To investigate mechanisms of hepatocyte injury, lymphocytes from 41 children with chronic liver disease were incubated with autologous liver cells in a microcytotoxicity assay. Cytotoxicity was significantly increased in 18 of 25 patients with chronic hepatitis B virus (HBV) infection, in five of nine with “autoimmune” chronic active hepatitis (CAH), and in only one of seven with histologically inactive liver disorders. There was a good correlation between cytotoxicity and biochemical and histologic markers of disease activity in children with autoimmune CAH, whereas in HB s Ag-positive disease a positive correlation was found only with serum alanine aminotransferase (SGPT). Children with autoimmune CAH receiving steroid treatment had normal cytotoxicity, whereas increased values were found in two of three HB s Ag-positive patients receiving prednisolone. Fractionation studies revealed that non-T cells were cytotoxic in both autoimmune and HB s Ag-positive chronic liver disease. T cell cytotoxicity was exclusively found in children with chronic HBV infection, particularly with HB e antigenemia. Blocking experiments showed that T-lymphocytes from HB s Ag-positive children reacted with HBV core antigen on the hepatocyte surface. Non-T cells were directed against hepatocyte membrane antigens in both HB s Ag-positive and Hb s Ag-negative children. These results suggest that different immune mechanisms of liver damage are involved in autoimmune and HB s Ag-positive chronic liver disease.

Ultrastructural changes were observed in 23 consecutive patients who died from fulminant hepatic failure due to hepatitis B virus (4 cases), sporadic non-A, non-B (7), or paracetamol (acetaminophen) overdose (12) and in 3 patients with subacute hepatic necrosis of unknown cause. The findings are described in detail in 12 of these patients. Fatal fulminant hepatitis was characterised by massive confluent necrosis accompanied by collapse of reticulin framework and sudden drop-out of liver cells. No aetiological distinction could be made between different viral causes of fulminant hepatitis on the basis of ultrastructural pathology. Parenchymal changes in viral cases varied from reversible non-specific necrosis to irreversible changes where fragmentation of endoplasmic reticulum, mitochondria and nuclei had occurred. Differences in ultrastructural pathology between non-viral (paracetamol overdose-induced) and viral fulminant hepatitis were apparent. Modification of endoplasmic reticulum with enlarged attached polyribosomes, breakdown of plasma membrane, accumulation of cytoplasmic amorphous material and karyorrhexis and karyolysis of nuclei were the most prominent features in non-viral cases.

In an observer variation study, six histopathologists independently coded 27 features and final diagnoses on the same 60 liver biopsies without knowledge of clinical data. Many features were shown by kappa statistics to have significantly high agreement, and differences were shown to be due in part to differences in definitions of particular features. Intraobserver variation was much lower than interobserver variation. Full agreement with respect to the final histopathologic diagnosis between the six observers occurred in only nine instances--all of these were cases of alcoholic liver disease. Some of the biopsies of acute hepatitis were confused with chronic active hepatitis, while others were mistaken for cholestasis, usually believed to have been related to drugs. Biopsies from patients with extrahepatic biliary obstruction were most commonly misdiagnosed as drug cholestasis. However, agreement was higher when the results were calculated in terms of observer pairs by comparing the final diagnosis of each observed with that obtained from each of the other five observers. The most striking finding here was the very high agreement rate (80%) between the two specialist pathologists compared with that obtained from the remaining fourteen pairs of observers who averaged an (43%) agreement rate.

Twenty-two patients with fulminant hepatic failure who deteriorated to grade-IV coma despite full supportive therapy were treated by repeated periods of haemoperfusion through columns containing activated charcoal. The procedure was well tolerated clinically. Eleven of the patients regained consciousness and ten left hospital. Follow-up liver biopsies in the first three patients at around six months after discharge from hospital showed restitution of the normal lobular architecture. Of the eleven treatment failures, haemorrhage was responsible for death in three, and in six brain herniation secondary to cerebral œdema was an important contributory factor. The column extracted most aminoacids from plasma, and during perfusion arterial concentrations of phenylalanine, tyrosine, and methionine-aminoacids known to be involved in the pathogenesis of the encephalopathy—fell significantly. The charcoal was coated with a biocompatible polymer, and there was no evidence for removal of coagulation factors. The extraction of platelets was below 30% in most instances, and in only two patients was there evidence that bleeding may have been precipitated by the haemoperfusion. These survival figures are to be compared with a previous survival figure of 10% in a series of ninety-two cases with fulminant hepatic failure and grade III or IV encephalopathy treated by full supportive measures.

HE ASSOCIATION between primary hepatic tumors T and androgen therapy was first reported in 1965,' and in a recent review we found 33 cases' of whom 14 had received androgens for Fanconi's aplastic anemia and the remaining 19 for various conditions including hypogonadism, transexualism, and non-Fanconi types of aplastic anemia. There has been much controversy concerning the malignant potential of these tumors and there are features of the tumors developing in the Fanconi patients which would appear to differentiate them from the remainder. Thus, tumors have been described as early as three months after starting androgen treatment3 and in one patient the presence of metastases was confirmed at po~tmortern.~ In contrast, in the non-Fanconi group, tumors seem to develop after prolonged androgen therapy (never less than three years), and although many have been described as hepatocellular carcinoma there are no instances of confirmed metastases. However, the follow-up of these cases has not always been adequate. We describe here three cases of hepatocellular tumor in non-Fanconi patients who had taken androgens for periods of 10-17 years and in whom both histologic, and long-term follow-up data are available.

The clinical, biochemical, and histological features of 27 children with syndromic paucity of the interlobular bile ducts are described. All presented in the first 5 months of life, 21 with jaundice, two with spontaneous bleeding due to vitamin K malabsorption in addition to jaundice, two with pruritus, and two with failure to thrive. Interlobular bile ducts were abundant in liver biopsies from five (18% of cases) in the first 6 months of life. The degree of portal fibrosis and cellular infiltrate was mild in all except three patients. Clinically significant heart lesions occurred in 52% but only 22% had peripheral pulmonary stenosis. Characteristic facial appearances were present in only 70%; embryotoxon and vertebral anomalies were present in 56 and 33%, respectively. Two infants died of cardiovascular complications, one of alimentary bleeding and one of progressive liver disease. Complications of vitamin K deficiency occurred in 15%, vitamin D deficiency in 30%, and vitamin E deficiency in 37%. Survivors at ages of 19 months to 16.5 years had considerable morbidity with pruritus occurring in 70%, jaundice in 48%, xanthomas in 30%, 74% having hepatomegaly and 63% splenomegaly. All had abnormal biochemical tests of liver function, 90% had growth retardation, and 50% developmental delay. We conclude that differentiation from extrahepatic biliary atresia can be difficult if biliary flow cannot be demonstrated. Prevention of fat-soluble vitamin deficiency is essential. Further research is required to decrease the morbidity associated with this syndrome in infancy.

We describe the clinicopathological findings in 12 cases of hepatic epithelioid haemangioendothelioma in order to identify diagnostic and prognostic features of this unusual vascular neoplasm. Three main tumour patterns were observed histologically: (1) a peripheral pattern with neoplastic cells scattered between fairly normal liver cell plates; (2) a cellular pattern showing a confusing admixture of pleomorphic tumour cells and atrophied hepatocytes set in a small amount of fibrous stroma; and (3) a scarred pattern with sparse tumour cells in a dense fibrous matrix. There were two types of vascular invasion: tuft-like intravascular proliferations of epithelioid cells and fibro-thrombotic venous occlusions. Awareness of these different aspects is important, as they are variably sampled by biopsy needles. The clinical course in this series was less favourable than that previously reported. Eight patients have died, in six instances of liver failure within 4-41 months of diagnosis. Extensive involvement of both lobes of the liver heralds imminent hepatic failure. The slow growth of metastases may justify liver transplantation in order to prolong life.

Parenchymal and non-parenchymal cells were harvested by washing the liver tissue core and needle after percutaneous biopsy. The cytological material obtained was suitable for morphological analysis, including showing the presence of surface and cytoplasmic antigens using labelled antibody techniques. This technique provides a combined cytological and histological approach to the diagnosis of liver disease.

Thirteen children (8 female) with primary sclerosing cholangitis are described, in whom the diagnosis was confirmed by the presence of characteristic changes on endoscopic retrograde cholangiopancreatography. Nine had clinical features of chronic inflammatory bowel disease 1 mo to 5 yr before the onset of primary sclerosing cholangitis (6 patients) or appearing simultaneously with primary sclerosing cholangitis (3 patients). In 4 patients clinical evidence of chronic inflammatory bowel disease was absent but 1 of the 4 was found to have microscopic colitis in colonoscopic biopsy specimens. Biopsies were not performed in the remaining 3 patients. High immunoglobulin G concentrations and positive antinuclear or smooth muscle antibodies were present in all patients except 1 who had been given immunosuppressants. In 7 patients treated with immunosuppressants and followed up for 9 mo to 10 yr there was modest symptomatic improvement. This improvement was accompanied by a fall in transaminase levels in 6 of the patients and histologic improvement in 3 of 4 patients who had undergone biopsy. Greater use of endoscopic retrograde cholangiopancreatography in the last 6 yr led to the identification of 10 of these 13 cases, suggesting a higher incidence of primary sclerosing cholangitis in childhood than would appear from the literature.

1. The degree of liver damage observed 48 h after administration of 14C ring-labelled paracetamol (3-23 mmol/kg) to rats was proportional to the amount of a highly reactive metabolite retained in the liver, bound covalently to hepatocellular proteins. 2. With increasing doses of paracetamol, urinary excretion of the glucuronide and sulphate conjugates reached a plateau, whereas the output of cysteine and mercapturic acid conjugates increased markedly. 3. The degree of covalent binding at 48 h was proportional to the rate of urinary elimination of these two latter conjugates in the first 24 h after dosing.

The zonal distribution of enzyme activities was measured by quantitative cytochemistry in cryosections of liver from three normal children and five infants with idiopathic hepatitis of infancy. Optimal conditions for cytochemical reactions were first validated in rat liver and subsequently used in human livers to quantify zonal activities of acid phosphatase (AP), succinate dehydrogenase (SDH), glutamate dehydrogenase (GDH), glucose-6-phosphatase (G6P) and NADPH-dehydrogenase (ND). In normal rat and human livers, activities were greater for SDH and G6P in periportal and for GDH and ND in perivenular hepatocytes, while AP was evenly distributed along the sinusoids. In five infants with idiopathic hepatitis of infancy (IHI), a similar trend of distribution was observed for the two mitochondrial (SDH and GDH) and the two microsomal (G6P and ND) enzymes, although the distribution gradient was less pronounced than, in normal livers. AP showed a mildly greater periportal than perivenular activity. This preliminary study shows that a similar metabolic zonation exists for these enzymes in human livers as is observed in rats.

The influence of ethanol on paracetamol-induced liver damage was studied in mice and related to changes in microsomal monooxygenases and plasma paracetamol metabolites in the same group of animals. Paracetamol (400 mg/kg body wt, p.o.) was administered alone or simultaneously with ethanol (3 g/kg, p.o.) to mice fed either a chow diet or pretreated for 4 weeks with a liquid diet containing ethanol (20% of energy). Acute ethanol administration protected against paracetamol hepatotoxicity, but this protection was complete only in mice not fed ethanol previously. Acute ethanol administration also appeared to reduce paracetamol monooxygenation in vivo, but ethanol (50 mM) added to microsomal incubations in vitro had no significant effect on paracetamol activation and covalent binding. The chronic ingestion of ethanol in the diet increased paracetamol-related liver damage, but there appeared to be no induction of paracetamol monooxygenation in these animals. We are unable to confirm current concepts that the potentiation of paracetamol hepatotoxicity by chronic ethanol ingestion and its reduction by acute ethanol administration result solely from contrasting effects of ethanol on cytochrome P-450, and alternative explanations are proposed.

Antibodies reacting with a human liver-specific membrane lipoprotein (LSP) have been detected using a sensitive and specific radioimmunoassay in 19 (51%) of 37 patients with primary biliary cirrhosis. The anti-LSP antibodies were found only in the later stages of the disease as judged by histological criteria, being present in 73% of those in stage IV, 44% of those in stage III, and none of those in stage I or II. Although there was no relationship between percentage binding and standard liver function tests, there was a close correlation between percentage binding of 125I-LSP by serum and the extent of piecemeal necrosis of periportal hepatocytes on liver biopsy. The timing of the anti-LSP response makes it very unlikely that it is involved in the pathogenesis of the early bile duct damage but the results of this and other studies suggest that antibodies to this hepatocyte membrane lipoprotein may be an important cause of periportal liver cell necrosis in both primary biliary cirrhosis and chronic active hepatitis and could be one of the factors determining progression to cirrhosis in both these conditions.

Of 47 consecutive infants with extrahepatic biliary atresia, effective bile drainage with the return of the serum bilirubin concentration to normal, was achieved in 17 (38%). Direct bile duct-to-bowl anastomosis, attempted in 15 infants, produced bile drainage in only those 4 (9%) in whom bile could be seen within the bile duct remnants at laparotomy. 13 (45%) of 29 infants subjected to portoenterostomy (direct liver-to-bowel anastomosis) had satisfactory prolonged bile drainage with normal serum bilirubin values. Although a correct preoperative diagnosis was made in each case, in 3 (6%) the 72-hour faecal rose bengal 131I excretion was greater than 10% of the injected dose, and in 5 (11%) the hepatic histology did not indicate bile duct obstruction, showing that both investigations are necessary for preoperative diagnosis. Preoperative clinical, laboratory, and hepatic histological features in the 16 jaundice-free survivors showed no significant difference when compared with the 31 infants in whom surgery was successful. Cholangitis occurred in only 7 (43%) of 16 infants with satisfactory bile drainage and was easily controlled with antibiotic treatment. No cutaneous enterostomies were performed. In most survivors liver function tests remain abnormal, but the patients are symptom-free. While it is too early to predict a long-term prognosis for these children, our eldest survivors are healthy and show normal development.

Summary Fourteen patients, aged 18 to 38 years, presented in the last trimester of pregnancy, or immediately post partum, with severe acute hepatic dysfunction. Liver biopsy confirmed the presence of acute fatty liver in six patients; the cause of hepatic dysfunction was presumed viral hepatitis in five patients, pre-eclampsia in two, and gram-negative septicaemia in one. There was one set of twins and the perinatal mortality was 33 per cent; there were three maternal deaths (21 per cent). Two survivors with acute fatty liver subsequently had successful pregnancies without evidence of hepatic dysfunction.

To test whether the increased prevalence of HLA-B8 reported in patients with alcoholic cirrhosis is due to the antigen being a genetic marker of susceptibility to liver damage from alcohol, patients who had cirrhosis of comparable clinical and histological severity were investigated for HLA-B8 status and cumulative alcohol intake. Both male and female cirrhotics with HLA-B8 had been drinking greater than 40 g alcohol/day for a shorter period of time (16.6 +/- 1.4 men, and 9.4 +/- 2.0 years, women) than their counterparts without this antigen (23.7 +/- 1.7, p less than 0.005, and 15.8 +/- 2.0 years, p less than 0.05, respectively), but the mean daily alcohol intake was similar whether patients had HLA-B8 or not. These results suggest that genetic determinants linked to HLA-B8 enhance the rate of development of liver damage in those who drink potentially hepatotoxic amounts of alcohol.

A study into the normal anatomy of the venous circulation of the gastroesophageal junction was undertaken using three complementary techniques (radiology, corrosion casting, and morphometry). Four distinct zones of venous drainage were defined as follows: (a) gastric zone, characterized by a longitudinal venous distribution; (b) palisade zone, composed of parallel vessels arranged in groups, lying mainly within the lamina propria; (c) perforating zone, characterized by "treble clef" shaped veins, which collect and channel blood into extrinsic veins; and (d) truncal zone, composed of four or five deep lying descending veins. This venous system appeared to be mainly distributed within the esophageal mucosal folds. The anatomic pattern suggests that venous flow is bidirectional at the palisade zone, which acts as a high-resistance watershed region between the portal and azygos systems. In patients with portal hypertension this normal vascular system has to accommodate greatly increased venous flow, and the anatomy as demonstrated here offers insight into variceal development.

To determine whether an autoimmune reaction to liver-specific proteins occurs in alcoholic liver disease, the cytotoxic effect of lymphocytes on isolated hepatocytes was determined in 27 alcoholic patients. Cytotoxicity was demonstrated in 15 of 17 patients with a histological diagnosis of alcoholic hepatitis, but in none of 10 with other forms of alcoholic liver disease. The ability of a liver-specific membrane lipoprotein to block the reaction suggests that sensitization to this antigen is responsible for the cytotoxicity. The demonstration of cytotoxicity using T cell-depleted, but not B and K cell-depleted, lymphocytes suggests an antibody-dependent cell-mediated reaction. There was a close correlation between cytotoxicity and the presence of alcoholic hyalin, liver cell necrosis, and piecemeal necrosis, but not with other histological features, immunoglobulins, autoantibodies, or standard liver function tests. The persistence of this autoimmune response, induced in some way by alcohol or one of its metabolites, may be important in the progression of acute alcoholic hepatitis to chronic liver disease.

Eleven cases of Wilson's disease presenting as fulminant hepatic failure were analysed retrospectively to determine the specificity or otherwise of the histological findings. All cases were cirrhotic, eight with a micronodular pattern. There was marked parenchymal collapse with ductular proliferation and mild inflammation. Other features included cholestasis, hepatocyte necrosis, microvesicular fat and nuclear vacuolation. Orcein staining demonstrated copper-associated protein in the periphery of cirrhotic nodules in all cases and also variably within nodules in eight cases. Copper was demonstrable by the rhodanine method in similar locations but the staining reaction was qualitatively weaker in all cases. Characteristically, there was staining of both parenchymal and mononuclear phagocytic cells. This triad of cirrhosis, strong copper-associated protein deposition and copper positivity was not present in a control group of 20 cases of fulminant hepatic failure of other aetiology and with a similar clinical presentation. It is concluded that in the clinical context of fulminant hepatitis the presence of cirrhosis should raise the suspicion of Wilson's disease and that, with routinely processed and stained tissue, including autopsy tissue, the diagnosis can be made histologically.

A striking, previously unreported pattern of orcein-positive deposits attributed to excess copper-binding protein was found in liver-biopsy specimens from twelve cases of Indian childhood cirrhosis. A comparable picture was found only in biopsy specimens from patients with Wilson's disease. Deposits seen in about 20% of the two hundred and seventy-nine cases in various control groups, which were associated with prolonged cholestasis, were slight in comparison. This finding could have important therapeutic and pathogenic implications.