Detection of hepatitis B surface antigen in fixed tissues of patients with cirrhosis and hepatoma

Summary HBsAg has been sought by light microscopy in liver specimens from patients with cirrhosis (79 cases) and hepatoma (99 cases). The study was carried out on fixed material using orcein staining, immunoperoxidase technique and indirect immunofluorescence. HBsAg was detected in the serum by radio-immunoassay (RIA) using Ausria II-125 in 38 patients with cirrhosis and in 36 with hepatoma. In the 38 seropositive cases of cirrhosis HBsAg-positive cells were observed in 31 (81.6%) by the orcein staining and in 32 (84.2%) by the peroxidase and immunofluorescence staining. Among the 36 seropositive patients with hepatoma, HBsAg was detected in the surrounding non-neoplastic part of the liver, cirrhotic or not, in 30 (83.3%) by orcein staining and in 34 (94.4%) by the immunoperoxidase method and immunofluorescence. Positive solitary-cells were seen occasionally in the tumor tissue in 16 cases using orcein, in 9 using peroxidase and in 7 by fluorescence, out of the 36 seropositive patients with hepatoma. The results of this study do not support the hypothesis of a direct oncogenic effect of HBsAg on the liver cells, since this antigen was detected mainly in the non-neoplastic part of the liver tissue and only occasionally in the tumor cells. Of the 63 cases of seronegative hepatoma, 3 showed some round orcein-positive inclusion bodies in the cytoplasm of the neoplastic and the non-neoplastic cells; these bodies were not stained by the two immunological methods.     

Aetiology of cirrhosis, hepatic fibrosis, and hepatocellular carcinoma.

Histological study of 69 cases of cirrhosis, 9 of severe generalised hepatic fibrosis, and 19 of hepatocellular carcinoma showed an association with alcohol, hepatitis B surface antigen (HBsAg) or a1-antitrypsin bodies in, respectively, 41 (cirrhosis), 5 (fibrosis), and 9 (carcinoma). Eight of the cirrhotic cases and two of the carcinoma cases had double associations, HBsAg being present in all. Torcein and aldehyde fuchsin staining gave both false positive and false negative results when compared with immunofluorescence and immunoperoxidase methods for HBsAg. Large amounts of copper were found in four cirrhotic livers, and moderate amounts in 13: the diagnostic value of copper staining is questioned.     

Nonimmunologic binding of horseradish peroxidase to hepatitis B surface antigen. A possible source of error in immunohistochemistry

In the process of establishing the specificity of direct immunoperoxidase staining of liver tissue for hepatitis B surface antigen (HBsAg), an affinity of free horseradish peroxidase (HRP) for HBsAg in hepatocytes (ground-glass cells) was found. Of 95 patients, the horseradish peroxidase reaction was only positive in the livers of the 35 who were chronically HBsAg seropositive and not in the livers from 60 control patients with alcoholic cirrhosis who were HBsAg seronegative. Comparison studies using the orcein technic and immunoperoxidase methods confirmed the observation that both free horseradish peroxidase (not conjugated to an antibody) and HRP conjugated to an antibody unrelated to HBsAg had an affinity to the cytoplasm of hepatocytes containing HBsAg. The precise nature of this affinity is not known, but it is probably due to a reaction between an activated carbohydrate moiety of horseradish peroxidase and the free amino group of HBsAG.     

Liver cell dysplasia and hepatocellular carcinoma: a histoiogical and immunohistochemical study

Liver cell dysplasia (LCD) was found in 28 (60%) of 47 patients with hepatocellular carcinoma (HCC); 22 (79%) of them had associated liver cirrhosis. LCD was more frequently observed in posthepatitic cirrhosis (82%) than in the other forms. Carcinoembryonic antigen (CEA), alpha-1-antitrypsin (AAT) and alpha-fetoprotein (AFP), as demonstrated by the peroxidase-antiperoxidase method, were similarly expressed both in normal and in dysplastic cells. Hepatitis B surface antigen was found in eight cases (17%), six of which were associated with LCD. HBsAg was rarely found in dysplastic cells and frequently displayed a peculiar perinuclear pattern. The possible preneoplastic role of LCD is stressed.     

Liver cell dysplasia and hepatitis B surface and core antigens in cirrhosis and hepatocellular carcinoma of autopsy cases.

Liver tissues of 180 autopsy cases of cirrhosis and hepatoma and 285 consecutive autopsy cases of other diseases were studied for liver cell dysplasia correlated with hepatitis B surface and core antigens (HBsAg and HBcAg) in liver cells and sera, and antibody to HBsAg (anti-HBs) in sera. Liver cell dysplasia was characteristic in cirrhotic livers, particularly with hepatoma. No significant difference was found in age and sex between cirrhotic cases with and without dysplasia. Rate of positive HBsAg in liver cells and sera was significantly high in cirrhotic cases with dysplasia with or without hepatoma. Massive pattern distribution of orcein-positive liver cells was statistically significant in cirrhotic livers with or without hepatoma, but morphological characteristics of orcein-positive liver cells could not be correlated in significance with dysplasia and hepatoma. HBcAg showed neither correlation with liver cell dysplasia nor hepatoma. It appears to correlate with active cirrhosis, marked liver cell degeneration and necrosis, and membranous diffuse type HBsAg in liver cells.     

LIVER CELL DYSPLASIA AND HEPATITIS B SURFACE AND CORE ANTIGENS IN CIRRHOSIS AND HEPATOCELLULAR CARCINOMA OF AUTOPSY CASES

Liver tissues of 180 autopsy cases of cirrhosis and hepatoma and 285 consecutive autopsy cases of other diseases were studied for liver cell dysplasia correlated with hepatitis B surface and core antigens (HBsAg and HBcAg) in liver cells and sera, and antibody to HBsAg (anti-HBs) in sera. Liver cell dysplasia was characterietic in cirrhotic livers, particularly with hepatoma. No significant difference was found in age and sex between cirrhotic cases with and without dysplasia. Rate of positive HBsAg in liver cells and sera was significantly high in cirrhotic cases with dysplasia with or without hepatoma. Massive pattern distribution of orcein-positive liver cells was statistically significant in cirrhotic livers with or without hepatoma, but morphological characteristics of orcein-positive liver cells could not be correlated in significance with dysplasia and hepatoma. HBcAg showed neither correlation with liver cell dysplasia nor hepatoma. It appears to correlate with active cirrhosis, marked liver cell degeneration and necrosis, and membranous diffuse type HBsAg in liver cells.     

Maternal serum alpha fetoprotein concentrations in mid trimester in hepatitis B surface antigen positive and negative subjects.

Maternal serum alpha fetoprotein (AFP) was measured as part of a routine antenatal screening programme in 48 patients positive for hepatitis B surface antigen (HBsAg). After exclusion of two cases with obvious obstetric abnormalities there was no difference in AFP concentrations between subjects who were positive or negative for HBsAg.     

Incidence in South-west Scotland of hepatitis B surface antigen in the liver of patients with hepatocellular carcinoma.

A retrospective examination in South-west Scotland of formalin-fixed paraffin-embedded liver tissue by an immunoperoxidase technique revealed hepatitis B surface antigen (HBsAg) in eight out of 81 cases (10%) of primary hepatocellular carcinoma (PHC) and in four out of 82 cases (5%) of cirrhosis. No positive staining was found in 112 controls without overt liver disease matched for age and sex. Unlike most previous studies showing an association between HBsAg and PHC, the present investigation was carried out in an area where HBs antigenaemia is infrequent and PHC is an uncommon tumour. While possibly hepatitis infection is an important cause of PHC, the association between HBsAg and PHC could be due merely to activation by the tumour of latent virus B in a previously infected person.     

Serological studies and ultrasonographic diagnosis of chronic liver disease in Kenya

Blood samples were obtained from blood donors and patients with chronic hepatitis (CH), liver cirrhosis (LC) and hepatocellular carcinoma (HCC) from provincial hospitals and Kenyatta National Hospital (KNH). Patients with chronic liver disease (CH, LC and HCC) underwent abdominal ultrasound screening as well. The blood samples were screened for Hepatitis B surface antigen (HBsAg), anti hepatitis C virus (Hep CV) antibodies and alpha fetoprotein (AFP). A total of 44665 blood samples from blood donors were screened for HBsAg between July 1991 and January 1993. Of these 4.1% were found to be HBsAg positive. A total of 983 samples were taken from chronic liver disease patients out of which 22.2% were found to be HBsAg positive. Sixty-three patients were found to have liver cirrhosis and 53 patients had HCC on ultrasound screening. Anti Hep CV antibodies were found in 4.3% (5/116) of patients with LC and HCC. AFP levels were found to be significantly higher in HCC patients than in LC patients, levels of above 200 ng/ml being diagnostic of HCC. Follow up of high risk patients, i.e. those with HBsAg positive, chronic liver disease, by ultrasound screening and AFP detection may be useful in the early detection of HCC.     

Patterns of hepatitis B surface antigen. Localization in cells of hepatocellular carcinoma.

Hepatitis B surface antigen (HBsAg) was identified with immunofluorescence, immunoperoxidase, and aldehyde fuchsin stains within tumor cells in three cases of hepatocellular carcinoma (HCC) from a series of liver biopsies from 172 consecutive cases of HCC. Two patterns of distribution and staining of HBsAg in cells of HCC were observed. In two of the three biopsy specimens, HBsAg was confined to solitary or small groups of tumor cells where a heavily stained inclusion occupied the entire cytoplasm displacing the nucleus. These inclusions corresponded to ground-glass cytoplasm with hematoxylin-eosin. The pattern is different in the other specimen where all the HCC cells in one area of the tumor showed a diffuse peripheral or perinuclear staining of the cytoplasm. In hematoxylin-eosin sections, these tumor cells showed partial transformation of the cytoplasm into the ground-glass appearance.     

Hepatitis B virus antigens in liver resections from patients with hepatocellular carcinoma

A high rate of hepatitis B virus (HBV) antigen carriers among patients with hepatocellular carcinoma (HCC) has been recorded from areas of endemic HBV infections in Africa and Asia, but there are only rare and contradictory data for Europe. 12 specimens of resected liver tissue were immunohistologically investigated for hepatitis B surface antigen (HBsAg) as well as for hepatitis B core antigen (HBcAg). HBsAg was contained in non-tumorous liver tissue in 66 per cent of these cases. In two cases detection of HBcAg in the liver provided evidence to replication of the virus. HBcAg plus HBsAg were present in tumour tissue in one case. All of the HBV antigen carriers did not have chronic hepatitis. On the other hand, all patients with chronic hepatitis had HBV antigens in their liver tissue. HBV antigens were detectable in 7 non-cirrhotic livers, but were contained in only one of two cirrhotic livers. These results are likely to suggest a possible relevance of HBV infection to the aetiology of HCC even in central Europe without customary liver cirrhosis.     

The histopathology of the liver in older patients with hepatitis B virus surface antigenaemia

Summary The histopathology of the liver and the detectability of intrahepatic hepatitis B virus (HBV) markers were studied in 34 autopsy cases in elderly patients (mean age 73.9 years, range 60–91 years) who had had a history of positive HBV surface antigenaemia prior to death. Seven of 14 persistent HBV carrier cases (group A) in which long-lasting HBV surface antigen (HBsAg) carriage in the sera had been confirmed by sequential assays, and 5 out of 15 HBV-infected people (group C, single assay) showed significant primary liver damages including chronic hepatitis, toxic hepatitis, liver cirrhosis and hepatocellular carcinoma. In 5 cases (group B), one of which was type B liver cirrhosis, HBsAg became negative and HBsAb appeared during the follow-up period (up to 33 months). Among confirmed HBV carriers, HBsAg and HBV core antigen were most frequently found in the liver of cirrhotic cases with and without hepatocellular carcinoma (5 of 6), whereas these were rarely detected in those with nonspecific changes or slight hepatitic activity (1 of 7). All 5 cases in group B were negative for histological HBV-related antigens and the findings in group C were variously interpreted. Post-mortem cases of the aged HBV carriers who survived their mean life expectancy represent an important population in which to study the natural history of HBV carriers.     

Hepatitis B surface antigen and hepatocellular carcinoma in Southern Africa

Summary Most series in Africa show a high percentage of hepatitis B surface Hepatitis antigen in hepatocellular carcinoma. Two groups of cases were investigated in this study. The one was derived from the autopsy material at Baragwanath hospital from subjects who had lived in Soweto, a large Black urban town. The second group consisted of male Black mineworkers generally originating from rural areas. A combination of the aldehydefuchsin stain and immunoperoxidase technique was used. The two groups showed totally different results. The Baragwanath series consisted of 24 hepatocellular carcinomas of which only 4 (17%) were HBsAg positive. Of the 24 cases, 14 had cirrhosis of which 9 were macronodular and 5 micronodular. Ten of these cases showed heavy iron overload. The series of male Black mineworkers comprised 22 cases of which 16 (72%) were HBsAg positive. Twelve of the 22 cases showed a macronodular cirrhosis and there were no micronodular cirrhoses. Only one case showed severe iron overload. These findings delineate two different populations of hepatocellular carcinoma in Southern Africa.     

p53 expression in hepatocellular carcinoma in a population in Singapore with endemic hepatitis B virus infection.

AIMS--To study the expression and clinical significance (if any) of p53 protein in hepatocellular carcinomas (HCC) arising in a population with endemic hepatitis B virus (HBV) infection. METHODS--Immunohistochemical staining was performed on formalin fixed, paraffin was embedded histological sections of 46 HCC cases using an antihuman p53 monoclonal antibody; serial sections were also stained for hepatitis B surface antigen (HBsAg), hepatitis B core antigen (HBcAg) and alpha fetoprotein (AFP). Nuclear p53 staining was assessed according to intensity (absent, weak or strong) and extent (< 5%, 6-25%, 26-50%, and > 50%) of positive cells. Tissue HBsAg, HBcAg and AFP were recorded as absent or present. RESULTS--The p53 protein was expressed in 35% (16 of 46) of HCCs; the positive rate in grade III/IV tumours (13 of 31; 42%) was higher than in grade I/II tumours (three of 15; 20%) but this was not statistically significant. HBsAg positive tumours showed almost the same proportion of p53 staining (11 of 29; 38%) as HBsAg negative ones (five of 17; 29%). CONCLUSIONS--The p53 protein was expressed in 35% of HCC cases. There was no statistically significant correlation between HBV infection and p53 protein expression. Similarly, there was no definite correlation between p53 positivity and tumour size, histological grade or vascular invasion.     

Pancreatic cancer and fibrinogen storage disease

BACKGROUND: Ductal adenocarcinoma is the most common type of pancreatic carcinoma while squamous, carcinosarcoma, sarcoma, giant cell carcinoma, and clear cell types are all rare. Hepatocellular fibrinogen storage disease is also an uncommon disorder which may be associated with hepatocellular carcinoma. Two cases of pancreatic carcinoma were encountered in a family with fibrinogen storage disease, further raising the possibility of a predilection to malignancy in this unusual disorder. The tumour in one case was of the rare clear cell type. These two cases are the basis for this report. METHODS: Sections were cut from retrieved paraffin embedded tissue and stained for routine histology. Immunohistochemistry using the avidin-biotin technique was applied for the expression of the markers p53 (D07), carcinoembryonic antigen (CEA), c-erbB-2, epithelial membrane antigen (EMA), and alpha-fetoprotein (AFP). RESULTS: Both cases were adenocarcinoma of pancreatic ductal origin. The tumour in one case showed features of a clear cell carcinoma. The tumour cells expressed p53, CEA, and EMA immunoreactivity and were negative for c-erbB-2 and AFP. CONCLUSIONS: Hepatocellular fibrinogen storage disease is rare and has been described in association with chronic hepatitis, cirrhosis, and rarely with hepatocellular carcinoma. This represents the first report of its association with carcinoma outside of the liver.     

Orcein-positive hepatitis B surface antigen and liver carcinoma: their association in an Eastern US population

We retrospectively examined the association of hepatitis B infection and hepatocellular carcinoma (HCC) in a US East Coast population using orcein staining of fixed liver tissue. Hepatitis B surface antigen (HBsAg) was present in non-neoplastic hepatocytes in eight of 53 cases of HCC, but in no cases of cholangiocarcinoma or metastatic tumor. In five of the eight positive cases, macronodular cirrhosis was present; in three positive cases, cirrhosis was absent. The rate of positivity in livers with both HCC and macronodular cirrhosis was 28%, compared with 4.7% in livers with macronodular cirrhosis but no carcinoma. The low, but significant association of HBsAg and HCC, both in the presence and absence of cirrhosis, suggests that HCC may develop in a subset of patients in the United States as a result of infection with hepatitis B virus.     

Immunohistochemical characterization of 130 cases of primary hepatic carcinomas

Primary liver carcinoma (PLC) may express a certain number of markers. Here we communicate results of an analysis of five such markers (alpha-1-antitrypsin--AAT--, carcino-embryonic antigen --CEA--, alpha-fetoprotein --AFP--, and superficial --HBsAg-- and core --HBcAg-- antigens of hepatitis B virus) by means of PAP techniques in 130 cases of PLC, comparing the neoplastic tissue and the non-tumorous liver. Three variants of PLC are distinguished: hepatocarcinoma (HC) (108 cases); cholangiocarcinoma (CC) (19 cases); and three cases of hepatocholangiocarcinoma (HCC). AAT was positive in 29 HC, 2 HCC, and negative in all 19 CC. CEA appeared positive in 16 HC, 16 CC and only one HCC. AFP was positive in two HC, and negative in all CC and HCC. HBsAg displayed positivity in 15 HC and one HCC, being negative in all 19 CC. HBcAg was positive in 4 HC, and negative in all CC and HCC. HBsAg was also positive in two neoplastic emboli associated with HC. On the non-tumorous liver tissue the immunohistochemical results showed positivity for AAT and CEA, but not for AFP. Therefore the present results confirm that in the geographical area from which these tumors proceed, PLC is closely correlated with HBsAg positivity and with cirrhosis.     

Hepatitis B virus antigens in primary hepatic carcinoma: immunofluorescent techniques on fixed liver tissue.

The presence of hepatitis B surface (HBsAg) and core (HBcAg) antigens was investigated by immunofluorescence in specimens of liver tissue obtained at necrospy in 107 patients with primary hepatic carcinoma. HBsAg was detected in the cytoplasm of liver cells in 16 cases, and in eight of them the antigen was also found in malignant cells. HBcAg, which was present in the nuclei of liver cells in eight cases, was detected in the nuclei of tumour cells in six of these and also in two other cases showing HBsAg, but not HBcAg, in the nonneoplastic tissue. Although most of the primary hepatic carcinomas studied were associated with cirrhotic changes in the non-neoplastic tissue, HBsAg and HBcAg were also detected in the absence of underlying cirrhosis. Hepatitis B virus markers were demonstrated in non-neoplastic tissue, mainly in patients with a well-differentiated carcinoma, and only in these cases were they found also in the neoplastic tissue. These results show that hepatitis B virus antigens, including HBcAg, can be detected in the neoplastic cells of well-differentiated carcinoma of the liver. Although these cells could have been infected after the malignant transformation, a direct oncogenic role of the virus cannot be excluded.     

Liver cell surface localization of hepatitis B antigen and of immunoglobulins in acute and chronic hepatitis and in liver cirrhosis.

This paper describes immunofluorescence studies on liver cell surface localization of hepatitis B surface antigen (HBsAg) and of IgG in acute and chronic hepatitis and in cirrhosis. In acute hepatitis B, HBsAg was found at the surface of hepatocytes in an early phase of the disease, but not during the recovery. This finding is consistent with the hypothesis that immune reactions to HBsAg may be responsible for the liver cell lysis. In HBsAg-positive chronic hepatitis and cirrhosis the antigen was found in the cytoplasm, but not on the surface of the hepatocytes, while in HBsAg-negative cases the antigen could not be detected in the liver cells. Both in HBsAg-positive and in HBsAg-negative chronic active hepatitis (CAH) and cryptogenic cirrhosis IgG bound to the membrane of the hepatocytes could be detected, suggesting a role of antibodies in the pathogenesis of the disease.     

Immunohistochemical Differentiation of Yolk Sac-type Alpha-Fetoprotein from Hepatic-type Alpha-Fetoprotein

In order to differentiate yolk sac type alpha fetoprotein (AFP) from hepatic-type AFP, 5 yolk sac tumors (YSTs) and 6 hepatocellular carcinomas (HCCs) were examined immunohistochemically by the peroxidase antiperoxidase (PAP) method for AFP, and paradoxical concanavalin A (P Con A) staining, which has been reported to detect glycoprotein including AFP. In all 5 YSTs, AFP was negative for P Con A staining. On the other hand, AFP was strongly positive for the same staining in all 6 HCCs. A similar staining pattern for AFP was observed in human yolk sac endodermal cells and embryonal hepatocytes. Thus, it was clarified that yolk sac type AFP was unable to bind with Con A, in contrast with hepatic-type AFP, on tissue sections. It was concluded that the PAP method for AFP and P Con A staining might facilitate the immunohistochemical differentiation of these two types of AFP, and that it would be us et ul for clarifying the histogenesis of various AFP secreting tumors.