A case of autoimmune hepatitis with a high titer of antimitochondrial antibody and normal gamma-globulinemia.

We report here a patient with chronic active hepatitis who had no markers for hepatitis viruses and no hyper-gamma-globulinemia, but had high titers of antimitochondrial antibody. Serum levels of alkaline phosphatase were normal, and antinuclear antibody, antismooth muscle antibody, and antiliver kidney microsome antibody tested negative. The titers of antimitochondrial antibody exceeded 1:640, and the positivity for anti-M2 was ascertained by using both ELISA and immunoblot with beef-heart mitochondria and a recombinant pyruvate dehydrogenase E2 subunit as antigens. This patient responded to ursodeoxycholic acid (UDCA) therapy in the beginning, but her hepatitis flared up during UDCA therapy. In contrast, she responded completely to corticosteroid therapy. The clinical course and histological findings of this patient strongly suggest that this patient has autoimmune hepatitis.

Immunological and virological predictors of outcome during interferon-alpha therapy of chronic hepatitis C.

Results from a multicentre, clinical trial of interferon-alpha2a (IFN-alpha2a) for the treatment of chronic hepatitis C are reported. Serum hepatitis C virus (HCV) RNA levels were monitored as follows: before, and 2 days after, the first administration of IFN-alpha2a; during and at the end of treatment; and 6 months after completion of therapy. Peripheral blood lymphocyte subpopulations were measured, by two-colour flow cytometry, before and 3 h after the first intramuscular (i.m.) administration of 9 mega units (MU) of IFN-alpha2a. Virological responders had a significantly lower pretreatment level of CD11+ CD8- lymphocytes. Biochemical responders had significantly lower pretreatment levels of CD11- CD8+, human leucocyte antigen (HLA) DR- CD4- and HLA DR- CD8+ populations, and a higher pretreatment HLA DR+ CD4- population. These pretreatment differences disappeared 3 h after the first i.m. administration of IFN-alpha2a. CD11- CD8+ and HLA DR+ CD8+ cell populations became significantly lower in virological responders 3 h after the first i. m. administration of IFN-alpha2a. HLA DR+ CD4+ cell populations were increased less in biochemical responders. Thus, T-lymphocyte subpopulations were different between responders and non-responders to IFN therapy and IFN-modulated host immunity. Multivariate analysis showed that the pretreatment CD11+ CD8- cell population was an independent predictive factor of response to therapy. On the other hand, patients whose serum HCV RNA cleared or decreased within the first 2 days of IFN-alpha2a therapy were more likely to achieve a virological response. This predictive factor, however, was not an independent factor by multivariate analysis. These results suggest that host immunity is an important factor in response to IFN therapy, and HCV clearance within the first 2 days is a good predictive factor of response.

Reduced in vitro immunoglobulin secretion from peripheral blood mononuclear cells in responders to high-dose interferon-alpha 2b treatment for chronic hepatitis C.

BACKGROUND/AIMS: Immunological status has been considered to correlate to the response to interferon therapy for chronic hepatitis C. The aim of this study was to evaluate the correlation between humoral immunity and long-term response to interferon treatment for chronic hepatitis C. METHODOLOGY: Seventy-one patients with chronic hepatitis C received 10 million units of interferon-alpha 2b three times a week for 24 weeks. Peripheral blood mononuclear cells were obtained before interferon-alpha 2b was administered and were cultured for 7 days. Immunoglobulin concentration in the culture supernatants was measured by enzyme-linked immunosorbent assay and correlation with the response to the therapy was evaluated. RESULTS: Serum ALT levels normalized in 51.4% and hepatitis C virus RNA disappeared in 35.7% six months after the end of therapy. Immunoglobulin production was significantly lower in the patients in whom serum ALT levels normalized than those in whom serum ALT levels remained elevated. The similar result was obtained when efficacy was evaluated on the basis of hepatitis C virus RNA disappearance. CONCLUSIONS: These results suggest that the less humoral immunity, the better response to interferon will be obtained in patients with chronic hepatitis C, meaning that the balance in T-helper function is one of key factors in the response to interferon treatment.

TT virus infection in patients with fulminant hepatic failure.

OBJECTIVE: A new DNA virus, which has been designated the TT virus, was discovered in 1997. It is not clear whether TT virus is a cause of any of the types of hepatitis. We conducted a case-control study to test the hypothesis that the presence of TT virus is a necessary condition for the development of fulminant hepatic failure in people who have non-A, -B, or -C hepatitis. METHODS: We studied 55 patients with fulminant hepatic failure [28 men, 27 women, mean (+/- SD) age, 47 +/- 15 yr], 32 patients with acute hepatitis (18 men, 14 women, mean age, 38 +/- 15 yr), and 200 healthy subjects (106 men, 94 women, mean age, 42 +/- 14 yr). TT virus DNA was detected in sera by a nested polymerase chain reaction using a primer set for genotype 1. RESULTS: TT virus was more frequently detected in patients with fulminant hepatic failure [in 33 of 55 (60%); 95% confidence interval (CI), 47-73%] than in those with acute hepatitis [in 8 of 32 (25%); 95% CI, 10-40%; p = 0.0016] or in healthy subjects [in 50 of 200 (25%); 95% CI, 19-31%; p < 0.0001]. TT virus was detected at a significantly higher rate in non-A, -B, or -C fulminant hepatic failure [in 18 of 22 (82%); 95% CI, 66-98%] than in fulminant hepatic failure of A, B, or C type [45%, 28-62%, 15/33; p = 0.007] or in non-A, -B, or -C acute hepatitis [24%, 3-44%, 4/17; p = 0.0003]. The logistic regression analysis selected TT virus (p = 0.0009), age (p = 0.0116), and etiology (p = 0.0309) as independent variables associated with fulminant hepatic failure (coefficient of determination, 0.2335). CONCLUSIONS: TT virus comparatively plays a role in the pathogenesis of non-A, -B, or -C fulminant hepatic failure.

[Positive rate of anti-mitochondrial antibody in Japanese corporate workers].

Anti-mitochondrial antibody(AMA) has been reported to be detectable in approximately 85% of patients with primary biliary cirrhosis(PBC). Therefore, a test for AMA is acceptable to be essential for diagnosing PBC. However, the positive rate in Japanese general population has not yet been determined. We tested sera from 1,145 corporate workers who took an annual health check and evaluated the liver of AMA-positive subjects. An indirect immunofluorescence method was used for screening AMA. ELISA and immunoblotting method were used for detecting anti-M2 in AMA-positive cases. AMA was detected in 5 of 1,145(0.44%) corporate workers. AMA positive rate was higher in females than in males(0.91% and 0%, respectively) and the AMA-positive people are all females over age 40. All of the AMA-positive sera are also positive for Anti-M2. Liver biopsy was performed in two AMA-positive cases and the histology was compatible with PBC in both cases.

Induction of hepatic stellate cell proliferation by LPS-stimulated peripheral blood mononuclear cells from patients with liver cirrhosis.

We studied hepatic stellate cell proliferation in vitro. Peripheral blood mononuclear cells (PBMC) from patients with chronic active hepatitis C (CAH) and liver cirrhosis (LC) were cultured for 24h in the presence or absence of Escherichia coli lipopolysaccharides (LPS). Hepatic stellate cell proliferation induced by the culture supernatants was measured, and interleukin-1 (IL-1) and IL-6 levels in the culture supernatants were quantified. Culture supernatants of LPS-stimulated PBMC from LC patients induced rat hepatic stellate cell proliferation by almost 2.8-fold (stimulation index, 2.83 +/- 1.41) compared with when the cells were cultured without addition of PBMC culture supernatants. Production of IL-1beta was significantly higher in the culture supernatants of both CAH and LC patients than in those of ten healthy controls (P < 0.01 and P < 0.05, respectively). But there was no significant correlation between IL-1 production and the induction of hepatic stellate cell proliferation by the culture supernatants. Although there were no significant differences in IL-6 production by LPS-stimulated PBMC among healthy controls and CAH and LC patients, we observed a significant correlation between IL-6 production and the induction of hepatic stellate cell proliferation in the culture supernatants of LC patients. Rat hepatic stellate cells themselves produced IL-6, and treatment with IL-6 antisense oligodeoxynucleotides suppressed the cell proliferation, suggesting that IL-6 is an autocrine growth factor for hepatic stellate cells. The addition of human recombinant IL-6 (hrIL-6) augmented rat hepatic stellate cell proliferation, indicating that excessive IL-6 may further facilitate cell proliferation. These findings suggest that a cytokine cascade including IL-6 may participate in hepatic stellate cell proliferation in LC patients when they are exposed to endotoxin.

Detection of anti-LKM-1(anti-CYP2D6) by an enzyme-linked immunosorbent assay in adult patients with chronic liver diseases.

Anti-liver kidney microsome-1 (LKM-1) autoantibody, which is a serological marker for autoimmune hepatitis type II, recognizes Cytochrome P450 IID6 (CYP2D6). This autoantibody is also detected in a portion of patients with chronic hepatitis C. Anti-LKM-1 has been measured by indirect immunofluorescence (IF) using rat liver and kidney sections. However, this method has some problems in specificity and is so laborious to handle with many samples. In this study, in order to determine anti-LKM-1, we established an enzyme-linked immunosorbent assay (ELISA) for anti-CYP2D6 using a recombinant CYP2D6 fusion protein. We studied sera from 29 patients positive for anti-LKM-1 by the new ELISA. We further studied sera from a total of 301 patients with various liver diseases and 100 sera from normal controls negative for anti-LKM-1 by the new ELISA. The specificity of the ELISA was ascertained by absorption tests using sera positive for anti-LKM-1. In 29 sera from patients positive for anti-LKM-1 by IF, we found a good correlation between the logarithms of the antibody titers determined by IF and ELISA indexes obtained by our new method. Anti-CYP2D6 was positive in 12 of 12 (100%) patient with autoimmune hepatitis type II and 16 of 17(94.1%) with chronic hepatitis C positive for anti-LKM-1 by IF. In other 401 sera negative for anti-LKM-1 by IF, anti-CYP2D6 was all negative except a few sera. We established a new ELISA for anti-LKM-1 (anti-CYP2D6). This ELISA system is sensitive, antigen-specific and easy to be done. Therefore, this assay allows a routine test of many serum samples, especially for diagnosing autoimmune hepatitis type II.

Differences in antigenic sites, recognized by anti-liver-kidney microsome-1 (LKM-1) autoantibody, between HCV-positive and HCV-negative sera in Japanese patients.

Anti-liver-kidney microsome-1 (LKM-1), which reacts with cytochrome P450 IID6 (CYP2D6), is an autoantibody present in autoimmune hepatitis type II, which affects primarily young patients. Recently, it has been shown some adult patients with chronic hepatitis C are also positive for anti-LKM-1. Thus, anti-LKM-1-positive patients can be classified into two subgroups: (1) those with autoimmune hepatitis type II and (2) those with chronic hepatitis C. We investigated the antigenic epitopes of CYP2D6 with which each of these two anti-LKM-1-positive subgroups reacted. Multiple deletion mutants of CYP2D6 were constructed from a human liver cDNA library and five recombinant fusion proteins expressed. Antigenic epitopes were determined by immunoblot analysis using these proteins. Anti-LKM-1 present in HCV-negative sera recognized at least two peptide regions of aa213-280 and aa341-477 of human CYP2D6. In contrast, anti-LKM-1 present in HCV-positive sera recognized only a single region of aa341-477. Thus, the sera of patients with autoimmune hepatitis type II and patients with chronic hepatitis C recognize different antigenic epitopes of the CYP2D6 molecule. To our knowledge, this is the first time LKM-1 autoantigens have been analyzed at the molecular level in Japanese patients.

Irregular regeneration of hepatocytes and risk of hepatocellular carcinoma in chronic hepatitis and cirrhosis with hepatitis-C-virus infection.

BACKGROUND: Hepatocellular carcinoma (HCC) commonly develops in patients with chronic hepatitis or cirrhosis of the liver caused by hepatitis-C-virus (HCV) infection. We prospectively studied whether irregular regeneration of hepatocytes is a risk factor for HCC in these patients. METHODS: 242 patients were enrolled after liver biopsy and followed up by ultrasonographic scanning every 3 months. We examined age, sex, platelet count, the diagnosis of cirrhosis or chronic hepatitis, liver-cell dysplasia, and irregular regeneration. We classified irregular regeneration as slight or severe, based on histological expression of pleiomorphism, anisocytosis, bulging, and map-like distribution of hepatocytes. FINDINGS: 37 of 63 patients with cirrhosis and 26 of 179 with chronic hepatitis were judged to have severe irregular regeneration. HCC was diagnosed in 33 of 63 patients with cirrhosis (29 had severe irregular regeneration) and 12 of 179 patients with chronic hepatitis (11 had severe irregular regeneration) during mean follow-up of 5.5 years (SD 4.1; range 1-16). Multivariate analysis with a proportional-hazards model showed severe irregular regeneration (relative risk 15.1 [95% CI 5.6-40.7], p<0.0001) and a diagnosis of cirrhosis (3.8 [1.7-8.2], p=0.0008) to be significant risk factors for HCC. Within the diagnostic categories, irregular regeneration was also significant (cirrhosis 6.8 [2.1-21.9], p=0.0014; chronic hepatitis 28.5 [2.9-276.4], p=0.0038). INTERPRETATION: We recommend that liver biopsy to look for irregular regeneration should be done in patients with HCV-related chronic liver diseases. Those with severe irregular regeneration should be followed up carefully.

[2-5 AS activity in serum and peripheral blood mononuclear cells for chronic active hepatitis C and the relationship to clinical outcome of interferon therapy].

Of 49 patients with chronic hepatitis C treated by interferon (IFN), we measured 2'-5'-oligoadenylate synthetase (2-5 AS) activity in peripheral blood mononuclear cells (PBMC) and serum before and after the IFN therapy and studied the correlation with the clinical outcome. Before IFN therapy, the levels of 2-5 AS in PBMC and serum were significantly higher in patients with chronic hepatitis C than in healthy controls, though there was no correlation between the 2-5 AS activity and the clinical outcome. When PBMC were stimulated with IFN in vitro, the induced 2-5 AS activities in patients with chronic hepatitis C were almost same as those in healthy controls. Among patients infected with hepatitis C virus (HCV) genotype II which was considered relatively resistant to IFN, patients whose HCV was disappeared from serum by IFN therapy showed good induction of 2-5 AS activity by IFN in vitro, whereas patients in which serum HCV remained positive after the therapy showed poor response to IFN in vitro. The levels of 2-5 AS in PBMC 2 months after IFN therapy were still higher in patients whose HCV was continuously disappeared from serum by the therapy (complete remission) than in healthy controls. The in vitro induction of 2-5 AS in patients whose HCV in serum remained positive after the therapy was significantly lower than in patients with complete remission. The induction of 2-5 AS activity in patients to whom IFN therapy was ineffective, significantly decreased after the IFN therapy as compared with the activity measured before the therapy. These findings suggest that measurement of 2-5 AS activity in PBMC in vitro through IFN therapy might be useful for predicting in vivo responsibility to IFN and also knowing the change of the responsibility.

Role of H-ras gene in chronic liver damage in mice. By using transgenic mice carrying a human C-H-ras proto-oncogene without mutations.

Hepatic tumors including hepatocellular carcinoma were generated by carbon tetrachloride in transgenic mice carrying a human c-H-ras gene (rasH2 mice). RasH2 mice express 2 to 3 times more ras protein (ras p21) in the liver than do non-Tg mice. When carbon tetrachloride was administered, the rasH2 mice produced about 5 times as many hepatic tumors than did the non-transgenic mice. However, neither the 10-100 times higher ras p21 expression required for murine fibroblast transformation by itself nor the mutational activation of the H-ras gene was observed in carbon tetrachloride-induced hepatic tumors. These results show that H-ras proto-oncogene expression in the murine liver, even if it is not high enough to transform cells, also causes liver tumors when CC1(4) are repeatedly given.

Possible association between HLA antigens and the response to interferon in Japanese patients with chronic hepatitis C.

Correlation between the major histocompatibility complex class I antigens (HLA-A, -B and -C) and the elimination from serum of hepatitis C virus in patients with chronic hepatitis C has not been understood. We analyzed HLA phenotypes and their relationship to the efficacy of interferon treatment. Of the 172 patients who were treated with 9 million units of interferon-alpha 2a three times a week for 6 months, 54 patients were responders and 118 patients were non-responders. No significant difference was observed between the 172 patients and 199 healthy subjects with regard to the frequencies of HLA-A, -B and -C antigen phenotypes. However, HLA-B55, B62, CW3 and CW4 frequencies were significantly higher in responders than in non-responders to the interferon treatment. CW4 was found to link with B62, but other phenotypes were independent each other. Patients with HLA B55, B62 and CW3 had a significantly lower viral load, and showed a better response to interferon. These results suggest that HLA system does not have an influence on the evolution towards chronicity of the disease due to hepatitis C virus, but HLA B55, B62 or CW4, and CW3 may be a virus quantity-regulating factors which then affect to response to the interferon treatment, indicating that these HLA antigens in conjunction with a viral peptide is a key target antigen for cytotoxic T lymphocytes in patients with chronic hepatitis C.

Susceptibility of experimental autoimmune hepatitis in transgenic mice overexpressing the c-H-ras gene.

Results from a recent study of ours have demonstrated the significant role of the wild-type ras gene in the development of hepatocellular carcinoma in rasH2 mice having prototype human c-H-ras genes. Chronic cell death and regeneration have been considered to work as co-carcinogens with wild-type ras gene overexpression in this model. To elucidate a role of gene overexpression in the occurrence of chronic inflammation, we tried to induce inflammation in the liver of rasH2 mice by immunizing them with the supernatant of a freshly prepared syngenic liver homogenate. Immunization resulted in a dense inflammatory infiltrate in the portal tract and focal necrosis with spots of fatty or foamy degeneration in the transgenic mouse liver; however, these observations were less frequently observed in non-transgenic mouse liver. Monocytes, granulocytes and plasma cell infiltration were observed in the livers of transgenic mice. An immunohistochemical study showed that CD3-positive lymphocytes also infiltrated the liver. The inflammatory infiltrate was still present in the transgenic liver 24 weeks after the last injection, but little infiltrate was observed at the same time in non-transgenic mice. No hepatic tumours could be produced over the 6 months duration of the study and the results are only preliminary. However, these results do suggest that overexpression of wild-type ras is partially responsible for the occurrence of autoimmune chronic hepatitis.

Keio multicenter trial in high-dose interferon-alpha 2b treatment for chronic hepatitis C. Keio Interferon-alpha 2b Study Group.

The efficacy of two different high-dose treatment of IFN-alpha 2b was evaluated in this study. Serum hepatitis C virus (HCV) RNA levels were semi-quantified by simplified reverse transcription-polymerase chain reaction. Seventy-one patients with chronic hepatitis C received 10 million units of IFN-alpha 2b daily for 2 weeks or twice a week for 24 weeks. Alanine aminotransferase (ALT) levels overall normalized in 78.1% and 51.6% of the cases at the end of the therapy and 6 months after that, respectively. HCV RNA disappeared in 71.9% and 35.7% of the patients at the end of the therapy and 6 months after that, respectively. There was no significant difference between the 2 different regimes. The efficacy of the treatment was fair in cases in which the pretreatment level of the viral amount was low. The results of this study indicate that daily administration of IFN in the first 2 weeks during 6-month course does not increase the efficacy of the therapy in such a high-dose treatment regime.

Cloning of variable regions of an antibody that reacts with the soluble fraction of human liver cells and its possible value in chronic liver disease.

A gene encoding the variable regions of the heavy and light chains of a mouse monoclonal antibody designated H2, which specifically reacts with human liver cells, was cloned into a phagemid vector. The clone of the variable region was designed to be expressed as a separate protein, the structure of which is the same as that of the mouse antibody. The cloned phage protein specifically reacted with anti-idiotypic antibodies produced in rabbits against the original mouse antibody, and this reaction was specifically blocked by the original antibody. The soluble protein, expressed as a fusion protein, was detected as a single 30-kd band on sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and specifically bound to an anti-H2 idiotypic antibody as determined by Western blot analysis. Sera of patients with various diseases were assayed for antibodies to anti-H2 by sandwich enzyme-linked immunosorbent assay (ELISA). Only sera from patients with chronic liver disease reacted strongly. This binding was specifically blocked by the cloned soluble protein. The nucleotide sequences of the variable regions were determined by the dideoxy chain-termination method, and the sequences were approximately 95% identical to those of other mouse immunoglobulins. These findings suggest that a human antibody with the same idiotype as a mouse monoclonal antibody that reacts with human liver cells, can be detected in patients with chronic liver disease, suggesting that autoimmunity may be partly responsible for these diseases.

Reduction of LAK-sensitivity and changes in antigen expression on hepatoma cells by sodium butyrate.

We demonstrated that sodium butyrate (SB) induced differentiation of functions in human hepatocellular carcinoma (HCC) cell lines. To investigate relationship between the sensitivity for cellular cytotoxicity and the cellular differentiation of HCC cells, the effect of SB on lymphokine-activated killer (LAK) sensitivity and antigen expression of a human HCC cells were studied. SB induced LAK-resistance of human HCC cell lines, HCC-T and HCC-M, time-dependently. A flowcytometric analysis of cell surface antigens revealed that SB markedly reduced the expression of laminin and fibronectin and increased the expression of liver-specific antigen defined by a mouse monoclonal antibody time-dependently, but did not modify that of major histocompatibility complex antigens, intercellular adhesion molecule (ICAM)-1, or CEA. Leukocyte function-associated antigen (LFA)-3 expression on HCC-T was reduced slightly by SB treatment. LAK sensitivity was inhibited by anti-laminin, but not with anti-beta 2-microglobulin, anti-HLA DR, anti-ICAM-1, anti-fibronectin, or anti-CEA. Anti-LFA-3 reduced LAK sensitivity of HCC-T, but not HCC-M, although the reduction was less than that obtained by anti-laminin treatment. These results provided evidence that SB induced LAK-resistance of human HCC cells according to cellular differentiation and extracellular matrix functionality played an important role in this LAK-mediated cell killing. Moreover, the structure expressed on HCC cells, which contributed to LAK cytolysis, was different for each HCC cell.

Assessment of histological features and outcome of interferon therapy in chronic hepatitis C.

The correlation between the histological features of liver biopsy specimens before interferon (IFN) treatment and the clinical effect of IFN administration on chronic hepatitis C was investigated. A study of the relation between several histological features that were graded in 60 liver biopsy specimens from chronic hepatitis C patients before IFN treatment disclosed that the grade of portal fibrosis was positively correlated with the grade of other inflammatory features, including piecemeal necrosis and portal and lobular inflammation. The degree of portal fibrosis adversely affected the rate of normalization of ALT levels in chronic hepatitis C during and after IFN treatment. We reexamined 36 liver biopsy specimens that showed a moderate degree of portal fibrosis, and found that the degree of piecemeal necrosis was inversely correlated with the extent of lymphoid follicle formation in the portal tracts. During IFN therapy, the group of chronic hepatitis C patients who showed marked piecemeal necrosis and less lymphoid follicle formation in the liver specimens had a poor response to IFN treatment, whereas another group that showed marked lymphoid follicle formation and little piecemeal necrosis in the liver specimens had a good response to IFN. These relationships gradually disappeared after the completion of IFN treatment.

Antibody-dependent cell-mediated cytotoxicity against cell lines generated by liver-specific idiotype-bearing antibody.

We produced a murine monoclonal antibody (mAb), designated H2-mAb, against a fractionated soluble phase of human liver homogenate which antibody reacted with human liver cells. A human antibody possessing the same idiotype as the H2-mAb, designated LSIA (liver-specific idiotype-bearing antibody), can be measured by a sandwich enzyme-linked immunosorbent assay, using the anti-H2 idiotype antibody. The serum level of LSIA in patients with histologically proven chronic hepatitis (CH) was significantly higher than that in healthy subjects and it was also higher than that in subjects with other diseases, including systemic lupus erythematosus. In a comparison between patients with CH type B and those with CH type C, there was no significant difference in serum levels of LSIA. It was possible to purify LSIA from the sera of patients with CH. The purified LSIA bound to the human cell lines Chang and HCC-M, derived from liver cells and a hepatoma respectively, but not to HeLa cells, a uterine carcinoma derivative. The reactivity of this mAb to HCC-M was weaker than that to Change. Moreover, the presence of LSIA caused an antibody-dependent cell-mediated cytotoxic challenge against Change cells in vitro.

Changes of antigen expression on human hepatoma cell lines caused by sodium butyrate, a differentiation inducer.

We have previously reported the effect of a differentiation inducer, sodium butyrate (SB), on human hepatocellular carcinoma (HCC) cell lines, demonstrating that it was a potent inducer of differentiation. In the present study, we investigated the alteration in expression of an antigen defined by a murine monoclonal antibody, H2, as well as alterations in the expression of other antigens, on the HCC cell lines HCC-T, HCC-M, and PLC/PRF/5, since it is known that specific antigenic changes occur during the differentiation of leukemic cells. The expression of the antigen defined by H2 increased immunocytochemically on HCC-T, HCC-M, and PLC/PRF/5 during treatment with SB. A flowcytometric study showed that almost all the HCC-T and HCC-M cells treated with SB highly expressed this antigen after 5 days' treatment. The antigen expression detected by H2 decreased after the removal of SB from the medium. On the other hand, antigen expression detected by another monoclonal antibody, 5C11, was not changed by this treatment. The expression of intracellular adhesion molecule (ICAM)-1 in HCC-T increased slightly, but that of beta 2-microglobulin and HLA-DR did not change. These results demonstrated that some antigen expression was changed by SB treatment and that the antigen defined by H2 seemed to be highly expressed on human HCC cells in the differentiated state.