A controlled trial of 6 months thrice weekly lymphoblastoid interferon versus no therapy in chronic hepatitis B virus infection. A preliminary analysis of the first 32 patients

A controlled trial of lymphoblastoid interferon versus no therapy in patients positive for HBsAg, HBeAg and DNA polymerase activity with separate randomisation for sexual preference and histology is underway. Thirty two patients have been followed for a minimum period of 6 months of whom 15 have been randomised to receive interferon thrice weekly for 6 months after a 5-day induction phase. Five treated patients developed an hepatitis-like illness during the 3rd month of therapy concurrent with an abrupt and complete loss of DNA polymerase activity from serum. In 3 this was permanent and anti-HBe subsequently developed; 2 of these have also lost HBsAg. In the other 2 patients inhibition of viral replication was transient. In 5 further treated patients DNA polymerase activity was completely inhibited throughout treatment only to return as soon as interferon was withdrawn. In this group serum aminotransferase became normal during treatment. In the remaining 5 treated patients, inhibition of DNA polymerase activity was never complete and serum aminotransferases were unaffected. All the control patients remain seropositive for HBsAg, HBeAg and DNA polymerase activity. The low seroconversion rate in treated patients and the absence of seroconversion in the control group are probably a reflection of the exclusion of patients with marked elevation of serum aminotransferases. The occurrence of an hepatitis-like illness in the 3rd month of therapy in a third of the patients and the loss of HBsAg in 2 of 3 who eventually seroconverted are likely to be a consequence of therapy rather than spontaneous events.(ABSTRACT TRUNCATED AT 250 WORDS)

Concanavalin A induced suppression of lymphocyte proliferation in chronic liver disease. A study of suppressor and responder populations in autologous and allogeneic systems

Abnormal T-cell regulation of lymphocyte proliferation may contribute towards tissue damaging mechanisms in chronic liver disease. We therefore studied Concanavalin A induced suppressor cell activity in T-T interaction in 47 patients with chronic liver disease, using both an autologous and an allogeneic system. In the autologous system, no differences were found between those with auto-immune chronic active hepatitis, HBsAg positive chronic active hepatitis, primary biliary cirrhosis, alcoholic liver disease and normal controls. However, several abnormalities were identified in allogeneic cultures with normal lymphocytes which allowed separate analysis of the influence of suppressor and responder cells from patients with chronic liver disease. An abnormality of the suppressor population was found in those with autoimmune chronic active hepatitis, primary biliary cirrhosis and alcoholic liver disease, while the responder population was abnormal in those with autoimmune or HBsAg positive CAH. Failure to demonstrate an abnormality in an autologous system may reflect a combined defect of suppressor and responder populations, and in this study the allogeneic system was a more sensitive index of abnormal cellular T-T interaction.

Altered natural history of hepatitis B in homosexual males--a reflection of altered immune responsiveness?

We suggest that subtle alterations in T-cell functions in male homosexuals makes them more likely to have silent anicteric hepatitis B infection, and perhaps more likely to become chronic carriers with significant chronic liver disease. This immunodeficiency might also explain why they respond less favourably to antiviral therapy. In future trials of antiviral therapy homosexual males should be randomised separately. Further studies of the immunomodulatory effects of new and existing agents will be of value in designing drug regimes particularly suited to the treatment of homosexual HBV carriers.

IgG production in 'autoimmune' chronic active hepatitis. Effect of prednisolone on T and B lymphocyte function

In vitro IgG production was measured using peripheral blood mononuclear cells from patients with autoimmune chronic active hepatitis (CAH) to determine whether the increased serum IgG levels were related to abnormalities of T or B lymphocyte function. A marked increase in spontaneous and pokeweed mitogen-induced proliferation of IgG producing cells was observed in 30 patients with untreated autoimmune CAH when compared with 25 normal subjects and 21 patients with autoimmune disease in whom a remission had been induced and maintained by prednisolone (P less than 0.01). Co-culture experiments clearly demonstrated that abnormalities of T lymphocyte function in untreated autoimmune CAH were responsible for the heightened IgG production in vitro. Pre-incubation of T lymphocytes from untreated patients with 5 X 10(-8)M prednisolone significantly reduced the number of cells producing IgG (P less than 0.05), suggesting that the modulation of the immune response following corticosteroid therapy is likely to be due to an alteration in T lymphocyte function.

In vitro studies on the mechanism of increased serum IgM levels in primary biliary cirrhosis

To evaluate the mechanisms underlying the increase in serum IgM in primary biliary cirrhosis (PBC) studies were designed to examine IgM production in vitro and to assess the relative contribution of intrinsic B cell activity and immunoregulatory T cell balance to IgM synthesis. The number of peripheral blood lymphocytes (PBL) producing IgM (spontaneous and pokeweed mitogen (PWM) stimulated) at the end of a seven day culture period was similar in PBC patients and control subjects while the amount of IgM synthesized (spontaneous and PWM stimulated) during this period was significantly greater in the patient group, implying that the amount of IgM produced per B cell was increased in PBC. Co-culture of autologous and allogeneic T and B lymphocytes and irradiation of T lymphocytes from patients and normal subjects clearly implicated abnormal suppressor T cell function, rather than autonomous B cell hyperactivity, as the cause of the increased IgM synthesis. Direct studies of T cell function indicated that although concanavalin A (Con A) activated suppressor cells inhibited proliferation of IgM producing B cells in the majority of PBC patients, they were unable to inhibit IgM synthesis. The demonstration of a disparity between IgM synthesis and the proliferation of IgM-producing B cells, together with the observation that the abnormality of T cell function is largely confined to the control of IgM secretion, is consistent with the presence of at least two different suppressor subpopulations regulating IgM production. In PBC the main suppressor cell abnormality seems to affect regulation of IgM secretion rather than B cell proliferation.

T suppressor cell function and number in children with liver disease

Con A stimulated suppressor cell function and the proportion of suppressor T cells were reduced in children with untreated chronic active hepatitis (CAH) but were normal in corticosteroid treated CAH patients, patients with severe acute hepatitis and inactive chronic liver disease. Adults with CAH also have defective suppressor function but a normal proportion of T suppressor cells. This difference may account for the observation that relapse after treatment withdrawal is less frequent in children than in adults.

Different mechanisms responsible for in vitro cell-mediated cytotoxicity to autologous hepatocytes in children with autoimmune and HBsAg-positive chronic liver disease

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 HBsAg-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 HBsAg-positive patients receiving prednisolone. Fractionation studies revealed that non-T cells were cytotoxic in both autoimmune and HBcAg-positive chronic liver disease. T cell cytotoxicity was exclusively found in children with chronic HBV infection, particularly with HBc antigenemia. Blocking experiments showed that T-lymphocytes from HBsAg-positive children reacted with HBV core antigen on the hepatocyte surface. Non-T cells were directed against hepatocyte membrane antigens in both HBsAg-positive and HbsAg-negative children. These results suggest that different immune mechanisms of liver damage are involved in autoimmune and HBsAg-positive chronic liver disease.

Immunoregulation of immunoglobulin production in normal infants and children

Proportion of T cell subsets, spontaneous and PWM stimulated immunoglobulin production by peripheral blood lymphocytes and concanavalin A- (Con A) stimulated suppressor cell activity on immunoglobulin production by B cells was studied in 37 infants and children, to investigate changes of these parameters with age. Proportion of suppressor/cytotoxic (T8+) T lymphocytes was significantly lower in children below the age of 5 years, while there was no difference in proportion of total T lymphocytes (T3+) and helper/inducer (T4+) T cells. Spontaneous production and secretion of IgG and IgM by lymphocytes from children of all age groups was similar to that found in adults, but lymphocytes of children below the age of 10 years showed a low response to PWM stimulation. The activity of Con A-induced suppressor cells in inhibiting B cells producing immunoglobulins was almost absent in infancy, gradually increased during childhood and reached adult levels in teenagers. A significant association between the proportion of T8+ cells and Con A-induced suppression of B cell proliferation and a relationship between T4+ cells and spontaneous Ig production indicated the increasing maturity with respect to both number and function of peripheral blood lymphocyte subsets with advancing age.

T lymphocyte sensitization to HBcAg and T cell-mediated unresponsiveness to HBsAg in hepatitis B virus-related chronic liver disease

Using a newly developed indirect T lymphocyte migration inhibition test, cell-mediated immunity to HBsAg and HBcAg was directly and simultaneously examined in a total of 21 patients with HBsAg-positive chronic liver disease (CLD), and in seven subjects whose sera contained anti-HBs (2 previous acute hepatitis B; 4 hepatitis B vaccine recipients and 1 chronic active hepatitis). T cell sensitization to HBcAg was invariably detected in the HBsAg-positive CLD patients tested (12/12), whereas T cell sensitization to HBsAg was not present in any of the patients (0/21). In contrast, T cell sensitization to HBsAg was present in all anti-HBs-positive subjects. These results support the hypothesis that the cellular immune response to HBcAg, rather than to HBsAg, is implicated in the pathogenesis of HBsAg-positive CLD. Moreover, the observation that the addition of T cells from patients with HBsAg-positive CLD to T cells from anti-HBs positive subjects in a ratio of 1 to 9 reversed their sensitization to HBsAg, suggests that a hyperactivity of HBsAg-specific suppressor T cell population may be responsible for persistent HBs antigenemia.

Mononuclear cell complement receptor blockade in primary biliary cirrhosis

Peripheral blood monocyte and lymphocyte receptors for Fc and C3b fragments were examined in vitro in patients with primary biliary cirrhosis and other chronic liver diseases using sheep red blood cells coated with anti-SRBC IgG1 (to detect Fc receptors) and with anti-SRBC IgM and complement (to detect C3b receptors). The number of C3b receptors detected on 100 monocytes was significantly lower in patients with primary biliary cirrhosis (23.0 +/- 12.0, mean +/- 1 SD) compared with normal controls (57.4 +/- 16.9) and other chronic liver disease (HBsAg negative chronic active hepatitis 62.0 +/- 17.0, alcoholic cirrhosis 50.9 +/- 4.0), while the number of Fc receptors detected on 100 monocytes was not significantly different in all the groups (primary biliary cirrhosis 72.8 +/- 28.6, chronic active hepatitis 74.7 +/- 14.0, alcoholic cirrhosis 58.0 +/- 13.5 and normal controls 69.6 +/- 19.9). When mononuclear cells isolated from normal individuals were pre-incubated with serum from patients with primary biliary cirrhosis before testing their receptor function there was a significant reduction in the number of C3b receptors detected per 100 monocytes (27.6 +/- 10.8) compared with pre-incubation with normal serum (72.0 +/- 18.0). This reduction in C3b-receptor function was again observed when the serum used for pre-incubation was depleted of circulating immune complexes; but when complement was further depleted from these sera, the number of C3b-receptors detected after pre-incubation was similar to normal values (64.0 +/- 11.8). Lymphocyte receptors showed a similar pattern of results. This implies a specific C3b receptor blockade on monocytes and lymphocytes from patients with primary biliary cirrhosis which appears to be because of blocking by serum factor(s) including complement fragments.

HLA A1-B8-DR3 and suppressor cell function in first-degree relatives of patients with autoimmune chronic active hepatitis

The relationship between suppressor T cell function and the inheritance of the A1, B8, DR3 haplotype was studied in 17 healthy, first-degree relatives of patients with autoimmune chronic active hepatitis. A marked defect in suppressor cell function was found significantly more often in A1, B8, DR3-positive relatives (5 of 7) compared with those who were A1, B8, DR3-negative (1 of 10; P = 0.017). Less marked abnormalities were also found in the A1, B8, DR3-negative relatives compared with A1, B8, DR3-negative control subjects. The results indicate that the defects in suppressor cell function in patients with untreated chronic active hepatitis result from the inheritance of genetic factors linked to the major histocompatibility complex and other gene loci and are not secondary to hepatic inflammation.

In vitro alpha-interferon treatment of peripheral blood mononuclear cells improves interleukin-2 activity in HBV-related chronic liver disease

To investigate mitogen-induced helper T cell activity in patients with HBV-related chronic liver disease (CLD), Interleukin-2 (IL-2) activity was assessed by an IL-2 bioassay using phytohaemagglutinin (PHA)-stimulated mononuclear cells (MNC). IL-2 activity was significantly reduced in patients with CLD (P less than 0.01), and was comparable to controls in those with minimal liver damage, indicating that decreased IL-2 activity is not due to the presence of HBV X MNC from 2 of the 3 patients treated with alpha-interferon (alpha-IFN) showed the highest IL-2 activity. In vitro preincubation of MNC with alpha-IFN before stimulation with PHA, led to a significant increase in IL-2 activity in all subjects (P less than 0.01). The improvement in IL-2 activity induced by alpha-IFN may be, in part, responsible for the therapeutic effect of this agent in HBV-related CLD.

Immune mechanisms in autoimmune chronic active hepatitis

Autoimmune chronic active hepatitis is a disease of unknown aetiology in which a dense mononuclear cell infiltrate in the portal areas of the liver is associated with ongoing necrosis of periportal hepatocytes. The finding of autoantibodies in serum, an increased frequently of HLA B8 DR3, a female predominance, an association with autoimmune diseases and the histological features all suggest a role for immunological reactions in the pathogenesis. Various immunological reactions have been demonstrated in vitro which could be of relevance to pathogenesis, including antibodies in serum directed against antigens expressed on the liver cell membrane, antibody-dependent cell-mediated cytotoxicity for autologous hepatocytes. T cell sensitisation to undefined hepatocyte antigen(s) and both antigen- and non-antigen-specific suppressor T cell defects. However, it is still unclear how these various phenomena interact in vivo and further studies are required to clarify their exact role in pathogenesis.

Antibodies to liver-specific lipoprotein in children with chronic liver disease due to "autoimmune" chronic active hepatitis, cystic fibrosis, and alpha 1-antitrypsin deficiency

Antibodies to a liver-specific lipoprotein complex (LSP) were determined by radioimmunoassay in the sera of 65 children with possible chronic liver disease. Thirteen had "autoimmune" chronic active hepatitis (CAH), 21 alpha 1-antitrypsin deficiency PiZ (alpha 1-ATD), all having significant liver disease, while 10 of 31 children with cystic fibrosis (CF) had abnormal biochemical tests of liver function, six having cirrhosis. Sera from 12 (92%) of 13 untreated CAH patients contained detectable anti-LSP antibodies, the highest titres occurring in those with anti-liver/kidney microsomal or smooth muscle antibodies. Titre of anti-LSP antibodies correlated with piecemeal necrosis of periportal hepatocytes in liver biopsies, but not with standard biochemical tests of liver function or serum immunoglobulin concentrations. The incidence of LSP antibodies fell as liver damage improved with immunosuppressant therapy, being positive in 18 of 26 sera from children in whom the transaminases were still above normal, but positive in only two of 20 in whom transaminase values were within the normal range (chi 2 = 16, p less than 0.001). Sera from two of 31 patients (6%) with CF contained anti-LSP antibodies as did sera from six of 21 patients with alpha 1-ATD. Antibody concentrations were lower than in CAH, and in alpha 1-ATD the presence of anti-LSP could not be correlated with the presence or absence of piecemeal necrosis on liver biopsy. Our data suggest that autoimmune mechanisms involving antibody to hepatocyte membrane components have a role in the pathogenesis of chronic liver disease in autoimmune CAH in children as in adults.(ABSTRACT TRUNCATED AT 250 WORDS)

A simple enzyme-linked immunosorbent assay for detection of anti-mitochondrial antibodies

An enzyme-linked immunosorbent assay (ELISA) has been developed for detection of anti-mitochondrial antibodies (AMA) in patients' sera. Results are expressed as a mitochondrial binding index, calculated from binding of patients' antibodies to rabbit liver mitochondrial fraction and total liver homogenate. The incidence of AMA detectable by this method in 22 sera from patients with primary biliary cirrhosis was identical with that obtained by conventional immunofluorescence testing (86%), and there was a highly significant correlation between titres determined by the two methods (r = 0.828; P less than 0.01) although ELISA titres were far higher. AMA were also detectable by ELISA in sera from 9/22 patients with chronic active hepatitis, 1/38 patients with other liver diseases, 3/16 patients with syphilis, 2/16 patients with non-hepatic autoimmune diseases and 0/29 normal blood donors; of these, only 2 chronic active hepatitis sera were positive for AMA by immunofluorescence. The ELISA titre was significantly correlated with the mitochondrial binding index determined at a serum dilution of 1:200 (r = 0.793; P less than 0.01), allowing an estimate of antibody titre to be made from ELISA screening at this single serum dilution. This assay, which is simple and reproducible, may be of value as an objective method of screening for AMA.

Lymphocyte cytotoxicity to autologous hepatocytes in alpha 1-antitrypsin deficiency

To investigate the possible role of cell mediated cytotoxicity in the pathogenesis or perpetuation of liver damage associated with alpha 1-antitrypsin (AAT) deficiency, peripheral blood lymphocytes (PBL) from 13 children with PiZZ phenotype who had had liver disease in infancy were incubated with autologous hepatocytes in a microcytotoxicity assay. There was a clear trend for cytotoxicity to increase with age and thus significantly increased cytotoxicity was found in four cases older than 2 years, while intermediate values were recorded in three children between 6 months and 2 years and normal values in children younger than 6 months. Fractionation of PBL into T and non-T-enriched subsets showed that both were involved in the cytotoxic reaction. A purified liver membrane lipoprotein preparation (LSP) inhibited both T and non-T-cell cytotoxicity suggesting that LSP is a major target antigen in this system. Control experiments performed in infants with neonatal hepatitis syndrome, but with normal Pi phenotype, showed consistently increased cytotoxicity values. Cell-mediated immune reactions directed against autologous hepatocytes develop late in the course of the liver disease associated with AAT deficiency. While this reaction cannot be involved in the pathogenesis of the initial liver lesion, it may contribute to perpetuation of liver damage.

Antigen specific suppressor cell function in autoimmune chronic active hepatitis

An indirect migration inhibition assay has been used to show that lymphocytes from 26 of 29 patients with autoimmune chronic active hepatitis (CAH) generated T lymphocyte migration inhibitory factors (T-LIF) in the presence of liver specific protein (LSP), compared with only 1 of 21 patients with HBsAg-positive chronic liver disease and none of 19 controls. Generation of T-LIF activity in response to LSP was not observed in any of 5 patients with autoimmune thyroid disease although their T lymphocytes did generate T-LIF activity in the presence of thyroid membrane antigens. T lymphocytes from 1 patient with autoimmune liver and thyroid disease generated T-LIF activity in the presence of both LSP and thyroid membrane antigens. The generation of T-LIF activity by T cells from autoimmune CAH patients was suppressed when these cells were co-cultured in a 9:1 ratio with T cells from normal subjects and patients with HBsAg-positive chronic liver disease, but was unaffected if co-cultured with T cells from other patients with autoimmune CAH. T cells from patients with autoimmune CAH did, however, suppress the generation of T-LIF activity by T lymphocytes from patients with autoimmune thyroid disease when these cells were cultured with thyroid membrane antigens. After pretreatment with cimetidine or mitomycin-C for 30 min, T cells from normal subjects lost their ability to inhibit the generation of T-LIF activity to T lymphocytes from autoimmune CAH patients. These results are consistent with the hypothesis that there exists a defect in the specific suppressor T cell population controlling the immune response to LSP in autoimmune CAH which is unaffected by disease activity and treatment and which may be of fundamental importance in the pathogenesis of the disease.

Controlled trial of a thymic hormone extract (Thymostimulin) in 'autoimmune' chronic active hepatitis

A randomised controlled trial of thymic hormone extracts (Thymostimulin) (1 mg/kg/day for seven days; 1 mg/kg/weekly thereafter) was undertaken in 30 patients (21 women, nine men) with treated, apparently inactive 'autoimmune' chronic active hepatitis during withdrawal of maintenance corticosteroid and azathioprine therapy. Reactivation of disease occurred in 26 patients (86%) during or after treatment withdrawal and was as frequent in the Thymostimulin treated (11 of 13; 84%) and untreated (15 of 17; 88%; p greater than 0.05) groups. Reactivation of disease was accompanied by a severe defect in concanavalin A induced suppressor cell activity, the magnitude of which was similar in the Thymostimulin treated and untreated groups (mean % suppression = 16.4 and 3.2 respectively; p greater than 0.05 vs 84.4 in control subjects). Further studies assessing the optimal dose, duration of treatment, and mode of administration are required to establish a therapeutic role for thymic hormone extracts in 'autoimmune' chronic active hepatitis.

Antibodies to alcohol altered liver cell determinants in patients with alcoholic liver disease

Circulating antibodies reacting specifically with hepatocytes isolated from ethanol pretreated rabbits have been demonstrated by two techniques - induced cytotoxicity and immunofluorescence. In the cytotoxicity assay antibodies were found in seven of 19 (39%) of patients with alcoholic fatty liver (with or without fibrosis), six of 13 (46%) of those with alcoholic hepatitis, 15 of 36 (43%) of those with cirrhosis, and seven of 14 patients (50%) of those with hepatitis and cirrhosis. In the immunofluorescence studies, nine of 15 sera induced a granular pattern of fluorescence on the ethanol pretreated hepatocytes; two sera which induced significant cytotoxicity did not induce immunofluorescence. No ethanol related antibodies were found in normal individuals or in patients with other types of acute or chronic liver disease. These results show that antibodies directed against ethanol altered liver cell determinants are present in the serum of 43% of patients with alcoholic liver disease, and suggest a mechanism whereby chronic alcohol consumption may, by inducing antigenic changes in hepatocyte membranes, trigger a cell damaging immune reaction.

Relationship between expression of hepatitis B virus antigens in isolated hepatocytes and autologous lymphocyte cytotoxicity in patients with chronic hepatitis B virus infection

Previous studies demonstrated that peripheral blood lymphocytes are cytotoxic to autologous hepatocytes in patients with chronic hepatitis B virus infection. We examined whether cytotoxicity is specifically directed against hepatocytes expressing HBsAg or HBcAg. Viral antigens were detected by immunofluorescence in isolated hepatocytes before and after exposure to T and non-T lymphocytes from 28 patients with chronic HBV infection in an autologous cytotoxicity assay. There was significant reduction in the percentage of HBcAg-positive hepatocytes after exposure to T lymphocytes and, to a lesser extent, after exposure to non-T cells; hepatocytes expressing HBsAg were not affected. Other studies showed that many hepatocytes containing HBcAg have IgG of anti-HBc specificity bound to their nuclei. In the present study, there was significant association between the presence of HBcAg and nuclear IgG in isolated hepatocytes. The ratio between the percentages of HBcAg-positive and IgG-positive hepatocytes (reflecting the proportion of hepatocytes containing free-core antigen) correlated significantly with reduction in HBcAg-positive liver cells after exposure to T cells, but not to non-T cells. These results suggest that hepatocytes in which viral replication is occurring and which express core determinants are susceptible to cell-mediated cytotoxicity and that T-cell cytolysis may be modulated by anti-HBc which is bound to hepatocytes.

Clinical and prognostic differences in fulminant hepatitis type A, B and non-A non-B

In 73 patients with fulminant viral hepatitis, non-A non-B hepatitis (NANB) was most common (43.8%), with hepatitis type A (HAV) diagnosed in 31.5% and hepatitis type B (HBV) in 24.7%. The non-A non-B group had a significantly longer duration from the onset of symptoms to the appearance of encephalopathy (median 21 days) compared with the HAV and HBV groups (medians 10 and seven days, p less than 0.01 and p less than 0.005 respectively). In the HAV group the severity of liver damage, judged by the maximum prolongation of the prothrombin time, was significantly less than in the HBV group (58 and 150 seconds prolonged respectively, p less than 0.005), and cerebral oedema was significantly less frequent (39% and 72% respectively, p less than 0.05). Consistent with this, the survival rate was higher in the HAV group (43.4%) compared with the HBV group (16.6%) and NANB group (9.3%) (p less than 0.005). These variations in presentation and clinical course may be a consequence of differences in the pathogenesis of the hepatic necrosis.

The value of clinical, biochemical, ultrasound and liver biopsy data in assessing patients with liver disease

To determine the value of clinical, biochemical, ultrasound and liver biopsy data in the management of patients with liver disease, eight doctors each assessed 75 case histories. With clinical and biochemical data alone, the predictive accuracy was significantly higher when identifying patients as 'medical' rather than 'surgical' (97 compared with 79%, p less than 0.001). However, when making a specific diagnosis as opposed to classifying into medical and surgical categories, clinical and biochemical information resulted in a much lower predictive accuracy for both medical (67%) and surgical (56%) patients. With ultrasound data the predictive accuracy increased to 91% when identifying patients as 'surgical'; with liver biopsy it increased to 99% when identifying patients as 'medical'. The value of the different data assessed involves more than an evaluation of diagnostic accuracy, and in this study the relative worth of each test was therefore assessed on a five point scale based on the effect of the information on the doctors. This included a willingness to give specific treatment and make a specific diagnosis, as well as classifying patients into medical and surgical categories and the confidence they felt in their diagnoses. After clinical, biochemical and ultrasound information the doctors were only prepared to give specific treatment to 11.9% of the medical and 9.3% of the surgical patients. After liver biopsy data, however, they were willing to give specific treatment to an additional 66.6% of the medical patients and 25% of the surgical patients. Further evidence of the value of liver biopsy information came from an analysis of the changes in the doctors' confidence in a diagnosis. Thus, 96 patients were assigned a correct specific diagnosis with clinical and biochemical data alone but none were considered as 'definitive' by the doctors; when liver biopsy information was provided 59 (61%) were placed in this category.