The dispersion-brightness relation for fast radio bursts from a wide-field survey.

Despite considerable efforts over the past decade, only 34 fast radio bursts-intense bursts of radio emission from beyond our Galaxy-have been reported1,2. Attempts to understand the population as a whole have been hindered by the highly heterogeneous nature of the searches, which have been conducted with telescopes of different sensitivities, at a range of radio frequencies, and in environments corrupted by different levels of radio-frequency interference from human activity. Searches have been further complicated by uncertain burst positions and brightnesses-a consequence of the transient nature of the sources and the poor angular resolution of the detecting instruments. The discovery of repeating bursts from one source3, and its subsequent localization4 to a dwarf galaxy at a distance of 3.7 billion light years, confirmed that the population of fast radio bursts is located at cosmological distances. However, the nature of the emission remains elusive. Here we report a well controlled, wide-field radio survey for these bursts. We found 20, none of which repeated during follow-up observations between 185-1,097 hours after the initial detections. The sample includes both the nearest and the most energetic bursts detected so far. The survey demonstrates that there is a relationship between burst dispersion and brightness and that the high-fluence bursts are the nearby analogues of the more distant events found in higher-sensitivity, narrower-field surveys5.

Cystic fibrosis liver disease: to transplant or not to transplant?

Biliary cirrhosis complicates some adults with cystic fibrosis (CF) and may require transplantation. Cardio-respiratory disease severity varies such that patients may require liver transplantation, heart/lung/liver (triple) grafts or may be too ill for any procedure. A 15-year experience of adults with CF-related liver disease referred for liver transplantation is presented with patient survival as outcome. Twelve patients were listed for triple grafting. Four died of respiratory disease after prolonged waits (4-171 weeks). Eight underwent transplantation (median wait 62 weeks); 5-year actuarial survival was 37.5%. Four died perioperatively; only one is alive at 8-years. Eighteen patients underwent liver transplant alone (median wait 7 weeks); 1- and 5-year actuarial survival rates were 100% and 69%. Three long-term survivors required further organ replacement (two heart/lung and one renal). Two others were turned down for heart/lung transplantation and four have significant renal impairment. Results for triple grafting were poor with unacceptable waiting times. Results for liver transplant alone were satisfactory, with acceptable waiting times and survival. However, further grafts were required and renal impairment was frequent. The policy of early liver transplantation for adults with CF with a view to subsequent heart/lung or renal transplantation needs assessment in the context of long-term outcome.

Predicting progressive hepatic fibrosis stage on subsequent liver biopsy in chronic hepatitis C virus infection.

Retrospective cross-sectional studies indicate that 20% with chronic hepatitis C virus (HCV) infection become cirrhotic within 20 years. Known risk factors for advanced hepatic fibrosis include age at time of infection, male sex, excess alcohol consumption and cytokine polymorphisms. Prospective study to assess and identify factors predictive of change in hepatic fibrosis stage in chronic HCV infection by interval protocol liver biopsy was performed. One hundred and five patients with paired liver biopsy specimens separated by a mean 41 months were recruited from a cohort of 823 HCV carriers. Five per cent developed worsening hepatic fibrosis by more than two stages. In 43% there was no change in fibrosis stage. Excessive alcohol intake currently (P = 0.037) or previously (P = 0.07) predicted progression. In contrast, always having a normal alanine transaminase (P = 0.038) and always being negative in serum for HCV RNA (P =0.067) predicted no progression. Three models were developed to predict outcome. Progressive fibrosis was predicted by baseline fibrosis (P = 0.018), steatosis (P = 0.02) and age (P = 0.017). The rate of progressive fibrosis was predicted by baseline fibrosis (P = 0.0002), steatosis (P =0.039) and lobular inflammation (P = 0.09). Fibrosis stage on the second biopsy was predicted by baseline fibrosis alone (P = 0.01). The rate of progression varies widely. Alcohol misuse is an important co-factor. Progressive fibrosis can be predicted at first liver biopsy, where baseline fibrosis is most critical, allowing targeted therapy for those with early disease and a significant risk of progression.

No association between hepatitis C and B-cell lymphoma.

Chronic viral infection has been implicated in the pathogenesis of B-cell lymphoma, and hepatitis C virus (HCV) infects lymphocytes. Chronic infection with HCV may result in B-cell proliferation. Individuals infected with hepatitis C are often co-infected with the RNA virus GB virus type C. Studies from Europe where hepatitis C infection is more common than in North America have shown a high prevalence of hepatitis C infection in patients with B-cell lymphoma. The aim of this study was to establish the prevalence of HCV and GBV-C infection in patients with B-cell lymphoma in an area of low HCV prevalence. One hundred patients with B-cell lymphoma (10 high grade, 46 intermediate grade, and 44 low grade) and 100 controls with nonhematological malignancies were studied. Serum was analyzed for HCV antibodies by third generation enzyme-linked immunosorbant assay, and HCV RNA and GBV-C RNA was analyzed by reverse transcriptase PCR. None of the controls or lymphoma patients had antibodies to HCV. HCV RNA was undetected in 60 out of 100 lymphoma patients tested. GBV-C RNA was detected in the serum of 5 out of 100 (5%) of lymphoma patients and in 3 out of 100 (3%) controls. Hepatitis C and GBV-C are, therefore, unlikely to play a major role in the pathogenesis of B-cell lymphoma in North America.

Enhanced ultrasensitive detection of structurally diverse antigens using a single immuno-PCR assay protocol.

Our studies of DNA damage and repair in autoimmune disease, lymphomagenesis, and carcinogenesis, require identification of an immunoassay approach that is capable of ultrasensitive detection in a routine human tissue biopsy of several physicochemically diverse antigens, some of which will be present at very low level. Immuno-polymerase chain reaction (immuno-PCR) is a recently described method for ultrasensitive antigen detection that combines the amplification power of PCR with a method similar to a standard antibody capture, enzyme-linked immunosorbent assay (ELISA). As a test of the universality of immuno-PCR, and as an assessment of the suitability of this method for our studies, we used a single immuno-PCR protocol to assay purified forms of the following physicochemically diverse antigens: oligomeric pyruvate dehydrogenase complex (PDC; Mr 8.5 x 10(6)), the promutagenic DNA base adduct O(6)-methylguanosine (Mr 298) and its monomeric repair enzyme, O(6)-methylguanine-DNA methyltransferase (MGMT; Mr 22,000), and a peptide from the N-terminus of MGMT (Mr 2310). We found that all antigens could be ultrasensitively assayed using the single immuno-PCR protocol. Assay limits observed using antigen-specific (primary) antibodies at 1 microg/ml, were in the approximate range of 10(2)-10(9) molecules, with O(6)-methylguanosine being detected most sensitively. Sensitivity of the antigen assay appeared to positively correlate with primary antibody titres determined by ELISA. Furthermore, we observed a substantial increase in detection sensitivity for all antigens by the use of primary antibodies at the higher level of 10 microg/ml. The latter approach permitted antigen assay within the approximate range of 10(0)-10(7) molecules. The combination of higher titre primary antibodies and their use at higher input level, produced an increase of immuno-PCR assay sensitivity of up to four orders of magnitude greater than those previously reported through the use of this assay to measure other antigens. This represents up to a nine order of magnitude increase in immunoassay sensitivity compared to ELISA. Our findings provide compelling evidence that immuno-PCR is indeed a universal ultrasensitive antigen detection method. Using the indicated assay enhancements. immuno-PCR performed as detailed here can offer greatly increased sensitivity for antigen measurement compared to other methods. Thus, our findings suggest that parallel quantitation of several different antigens in very small samples of human tissue will be readily attainable using immuno-PCR.

Donor factor V Leiden mutation and vascular thrombosis following liver transplantation.

The most commonly detected hypercoagulable state involves an abnormal factor V protein synthesized by the liver in which arginine at position 506 is replaced by glutamine as a result of a single-point mutation in the factor V gene (factor V Leiden). Liver transplantation is complicated by hepatic vascular thrombosis in up to 15% of cases, resulting in graft loss in most instances. This retrospective study examined the effect of the factor V Leiden mutation on the risk of hepatic vessel thrombosis after liver transplantation. The mutation was sought by polymerase chain reaction and Mnl I digestion of DNA where available from 214 recipients and 276 donors receiving 319 liver transplants. No donors or patients were homozygous for the factor V Leiden mutation. The prevalence of the heterozygous mutation was 19 of 276 (6.9%) in donors and 19 of 214 (8.9%) in recipients. Forty-one thrombotic episodes occurred after transplantation in the 276 transplants in which donor DNA was available for analysis; 22 involved the hepatic artery, 9 involved the portal vein, and 10 were deep venous thromboses. A donor factor V Leiden mutation was detected in the donor in 6 of 41 (14.6%) with any thrombotic event compared with 13 of 235 (5.5%) without (P = 0.03). The relative risk of any thrombosis with this mutation was therefore 2.32 (95% confidence interval [CI], 1.12-4.81). The factor V Leiden mutation was present in the donor in 4 of 31 (12.9%) cases complicated by hepatic vessel thrombosis (which always led to graft loss or death) and 15 of 245 (6.1%) cases without (P = 0.16). The relative risk of hepatic vessel thrombosis in the presence of this allele was therefore 2.00 (95% CI, 0.78-5.14). As anticipated, the presence of this allele in the recipient was not associated with deep venous or hepatic vessel thrombosis. The factor V Leiden mutation in the donor liver is not a major risk factor for hepatic vessel thrombosis and subsequent graft loss after liver transplantation.

Repair of DNA lesion O6-methylguanine in hepatocellular carcinogenesis.

Evidence from both experimental carcinogenesis and studies in human cirrhotic liver suggest that defective repair of the promutagenic DNA base lesion, O6-methylguanine, is a factor in the multistep process of hepatocellular carcinogenesis. Ubiquitous environmental alkylating agents such as N-nitroso compounds can produce O6-methylguanine in cellular DNA. Unrepaired, O6-methylguanine can lead to the formation of G --> A transition mutations, a known mechanism of human oncogene activation and tumour suppressor gene inactivation. Combined treatment of rodents with an agent producing O6-methylguanine in DNA, and an agent promoting cell proliferation, leads to development of hepatic nodules and hepatocellular carcinoma (HCC), cell division, hence DNA replication, being required for the propagation of tumorigenic mutation(s) in hepatocyte DNA. The paramount importance of O6-methylguanine in hepatocellular carcinogenesis is indicated by the observation that transgenic mice engineered to have increased hepatic levels of repair enzyme O6-methylguanine-DNA methyltransferase (MGMT) are significantly less prone to hepatocellular carcinogenesis following alkylating agent treatment. Cirrhosis is a universal risk factor for development of human HCC, and a condition that is characterized by increased hepatocyte proliferation as a result of tissue regeneration. Levels of the human repairing enzyme for O6-methylguanine were found to be significantly lower in cirrhotic liver than in normal tissue. In accord with findings from animal models, this suggested a mechanism in which persistence of O6-methylguanine due to defective DNA repair by MGMT, together with increased hepatocyte proliferation, might lead to specific gene mutation(s) and hepatocellular carcinogenesis. Screening for the presence and persistence of O6-methylguanine in human DNA presently involves formidable technical difficulty. Indications are that such limitations might be overcome by the use of an ultrasensitive method such as immuno-polymerase chain reaction (PCR). This approach should allow parallel measurement of DNA adduct and repair enzyme in routine liver biopsy samples. It might also enable investigation of O6-methylguanine in human genes specifically associated with hepatocellular carcinogenesis. Given the wide variation in human MGMT levels observed between individuals, tissues, and cells, this technology should be adapted to permit the ultrasensitive localisation and measurement of adducts and repairing enzyme in liver biopsy tissue sections. Ability to ultrasensitively measure O6-methylguanine, and its repair enzyme, should prove valuable in the risk assessment of cirrhotic patients for developing hepatocellular carcinoma.

Hepatocellular carcinoma in primary biliary cirrhosis and its impact on outcomes.

In this study we have determined the incidence of hepatocellular carcinoma (HCC) development in primary biliary cirrhosis (PBC) and its effects on patient survival. Six hundred and sixty seven patients with liver histology compatible with or diagnostic of PBC were seen over a 20-year period. Two hundred and seventy three patients who had stage III or IV disease on their last biopsy and who had been followed up for at least 1 year following that biopsy (total follow-up with advanced disease 2,010 patient years) were identified (243 female, 30 male). Patients who developed HCC were identified and their confounding risk factors were excluded. Mayo risk scores were calculated for each clinic attendance and expected survival for each time point was compared with subsequent actual survival. Sixteen cases of HCC were seen in the patients with stage III or IV disease on last biopsy, providing an overall incidence of 5.9% in this group. Fourteen of these patients had died of HCC related causes, and 2 patients were alive at the census point. The incidence of HCC was significantly higher in males with stage III/IV disease than in females (20% vs. 4.1%, P < .005). Nine of one hundred and eight (8.3%) total female deaths in this group was attributable to HCC compared with 5 of 11 (45.5%, P < .05) male deaths. HCC was not seen in any of the 394 patients with stage I and II PBC followed-up over the same time period. Throughout the disease course of all PBC patients with HCC, the Mayo prognostic model over-predicted survival. Whereas it is a relatively rare complication of cirrhotic PBC in women, HCC is a relatively common cause of death in male PBC patients with cirrhosis. HCC typically develops several years after the onset of cirrhosis, and is poorly predicted by prognostic models. In view of these findings, consideration should be given to careful screening for HCC in male PBC patients with cirrhosis. The risk of HCC development may be an additional reason to consider earlier transplantation in these patients.

Characterisation of the DNA repair enzyme for O(6)-methylguanine in cirrhosis.

BACKGROUND/AIMS: The reason(s) why cirrhosis is a major risk factor for the development of hepatocellular carcinoma worldwide are largely unknown. In a previous preliminary study, we reported deficiency of the repair enzyme for the highly promutagenic and potentially carcinogenic DNA base lesion, O(6)-methylguanine, in cirrhotic human liver. The aims of the present study were: (i) to confirm this observation in an extended series of cirrhosis patients, using a new DNA repair enzyme assay approach, (ii) to characterise the enzyme, in particular to seek physicochemical differences between control and cirrhotic liver that might account for the enzyme defect in cirrhosis, and (iii) to examine the relationship between magnitude of DNA repair deficiency in cirrhotic liver and aetiology of cirrhosis or male sex, both of which are independent risk factors of hepatocellular carcinoma. METHODS: Tissue extracts containing DNA repair enzyme, O(6)-methylguanine-DNA methyltransferase, were prepared from liver biopsy samples from 41 patients with cirrhosis (10 viral (HBV and HCV), 17 alcohol and 14 autoimmune), 17 patients with non-cirrhotic diseased liver and 4 patients with histologically normal liver. Repair enzyme levels and electrophoretic profiles were both obtained using sodium dodecyl sulphate-polyacrylamide gel electrophoresis in conjunction with fluorography and densitometry. RESULTS: We have demonstrated the feasibility of both analysing and reproducibly measuring DNA repair enzyme in small liver biopsy samples using the gel electrophoresis/densitometry approach for enzyme assay. Using the new densitometric assay approach, levels of O(6)-methylguanine-DNA methyltransferase were found to be significantly lower in cirrhotic compared to non-cirrhotic liver, with very low values obtained for two individuals who were incidentally found to have small hepatocellular carcinomas at the time of liver transplantation. There were no significant differences in enzyme levels between patients with cirrhosis of different aetiology, and between male and female patients. Methyltransferase was found to be present in all liver extracts as a major 23.1 kDA protein, along with other less abundant enzyme forms of both slightly higher and lower molecular weight; there were no obvious differences in size and relative abundance of these enzyme forms between liver samples from any of the patient and control groups. Our findings of enzyme instability and multiple forms led us to examine methyltransferase's amino acid sequence for the presence of primary structure motifs indicative of targeted degradation of the enzyme by intracellular proteases. We have identified a eukaryotic thiol protease active site motif and two cyclin-like "destruction box" motifs in the N-terminal half of the enzyme. CONCLUSIONS: Our findings provide further evidence suggesting that deficiency in the ability to repair O(6)-methylguanine-DNA underlies the increased risk of hepatocellular carcinoma seen in patients with cirrhosis. The degree of DNA repair deficiency not varying between cirrhotic patients subgrouped according to other risk factors for hepatocellular carcinoma, suggests that DNA repair deficiency is only one of several contributory factors in liver carcinogenesis. The presence in methyltransferase of thiol protease active site and "destruction box" motifs may be of relevance to observed repair enzyme instability and presence of multiple enzyme forms in human liver extracts.

Immunolocalization of pS2, a putative growth factor, in pancreatic carcinoma.

pS2 is a 60 amino acid secretory polypeptide which belongs to a newly described family of trefoil-shaped growth factors. It is widely distributed throughout the gastrointestinal tract, particularly adjacent to damaged mucosa, and is also expressed by some epithelial tumours such as breast carcinoma. The aim of this study was to examine the expression of pS2 in pancreatic cancer. The presence of pS2 was analysed immunohistochemically using two antibodies, a polyclonal (pNR-2) and a monoclonal (pS2TM) in 42 cases of pancreatic adenocarcinoma and 10 cases of ampullary carcinoma. The findings were compared with chronic pancreatitis and normal pancreas. No immunostaining was seen in normal pancreas, with the exception of one area of ductular proliferation, and although 8/10 cases of chronic pancreatitis expressed pS2, it was focal and confined to the occasional duct. In contrast, a significant proportion of malignant cells in 23/42 (55%) of pancreatic adenocarcinoma and 8/10 (80%) of ampullary tumours expressed immunoreactive pS2. The finding of pS2 expression in more than 50% of pancreatic and ampullary carcinomas in contrast to the findings seen in chronic pancreatitis and normal pancreas suggests that pS2 may play an important role in the growth of these highly malignant tumours.

Twenty-seven years of liver transplantation at Addenbrooke's Hospital, Cambridge.

Twenty-seven years ago, liver transplantation at Addenbrooke's Hospital was a very experimental procedure. It is now recognized as an accepted mode of treatment for end-stage liver failure. In the coming year, over 600 liver transplants will take place in the United Kingdom. All western European countries now have an active liver transplantation program and more than 17,000 liver transplants have already been performed. As results improve and the procedure becomes more readily and widely accepted, the donor shortage is likely to get worse, and will be only partially met by the introduction of split livers and living-related donors.

Expression of glutathione S-transferases in normal and malignant pancreas: an immunohistochemical study.

The glutathione S-transferases (GSTs) are a family of detoxification and metabolising enzymes, which have been linked with the susceptibility of tissues to environmental carcinogens and resistance of tumours to chemotherapy. Environmental carcinogens have been implicated in the pathogenesis of pancreatic carcinoma, which is also a tumour characterised by marked chemotherapeutic drug resistance. In this study 26 pancreatic adenocarcinoma and 12 normal pancreatic samples were examined immunohistochemically for expression of pi (acidic), alpha (basic), and mu (neutral) GST. Fourteen (54%) of the tumours expressed pi GST alone, two (8%) expressed both pi and alpha GST, and two (8%) showed immunoreactivity with alpha GST alone. In the normal pancreas the intralobular ducts and centroacinar cells expressed pi GST alone whereas the large ducts expressed both pi and alpha GST. The acinar cells showed immunoreactivity only with anti-alpha GST. Mu GST was not expressed by normal or malignant pancreas. Expression of pi GST by pancreatic carcinoma may be a marker of the malignant phenotype and be induced during neoplastic transformation. Alternatively it could possibly reflect cell of origin, suggesting that the tumour arises from the centroacinar cells or intralobular ducts, or both rather than the large ducts.

Clinical features and prognosis of hepatocellular carcinoma in Britain in relation to age.

A retrospective review of 110 cases of hepatocellular carcinoma (HCC) from the North East of England was undertaken to examine clinical features and prognosis in the 52 patients 65 years of age or more at presentation in comparison with 58 patients under 65. Symptoms and signs at the time of presentation were similar in the two age groups. Only 25% of patients were known to be cirrhotic at the time of presentation, but this increased to 75% after investigation. Although tumour stage at diagnosis was no worse in the elderly group, these patients received less intensive investigation and were significantly more likely to receive conservative therapy (p = 0.03). Although median survival was significantly worse in those aged > or = 65 years than in those < 65 years (10.5 and 18.5 weeks respectively; p = 0.02) this adverse effect disappeared when adjusted for the effects of treatment.

Expression of mutant p53 protein in hepatocellular carcinoma.

p53 mutations are a common genetic finding in hepatocellular carcinoma from areas of high aflatoxin exposure. Recent small studies have shown that p53 gene mutations may be less common in areas with a low prevalence of hepatocellular carcinoma such as Great Britain. The protein product of mutant p53 can be detected immunohistochemically because of its longer half life in comparison with native protein. This study used a novel monoclonal antibody DO-7, raised against recombinant p53 and effective in routinely processed biopsy specimen tissue, to detect the mutant protein in a series of 45 cases of hepatocellular carcinoma occurring in white subjects resident in the United Kingdom. Focal nuclear labelling was seen in four cases (9%); surrounding cirrhotic tissue in one of these was negative for p53 expression. This study shows that p53 mutations are a rare event in hepatocarcinogenesis in Great Britain, an area of low aflatoxin exposure, and supports the concept of geographical variations in the cause and pathogenesis of hepatocellular carcinoma.

Expression of transforming growth factor alpha in human hepatocellular carcinoma.

Transforming growth factor alpha (TGF-alpha) is a mitogenic polypeptide which acts on the epidermal growth factor receptor (EGFR). The aim of this study was to examine the expression of TGF-alpha in human hepatocellular carcinoma (HCC) and surrounding cirrhotic tissue, and to compare it with normal liver. Immunoreactive TGF-alpha was detected using two antibodies raised against its C terminus, a polyclonal antibody 26T and a monoclonal antibody Ab-2. In normal liver immunoreactive TGF-alpha was localised strongly to bile duct epithelium and weakly in occasional parenchymal cells but was notably absent from perisinusoidal and Kupffer cells. Eight out of twenty-eight (28%) cases of HCC expressed TGF-alpha as demonstrated by cytoplasmic staining with both antibodies and in four cases additional membrane immunoreactivity was demonstrated using 26T. However, where cirrhotic tissue surrounding TGF-alpha positive tumours was available for analysis immunoreactive TGF-alpha was detected in only 1/7 cases. TGF-alpha synthesis by malignant hepatocytes was supported by the detection of specific RNA by Northern blotting from two cases with TGF-alpha immunoreactivity. These results implicate bile duct epithelium as an important source of TGF-alpha in human liver. Furthermore, in HCC the expression of TGF-alpha in some cases, together with paucity of TGF-alpha immunoreactivity in surrounding cirrhotic tissue, suggests that TGF-alpha may play a role in continued cell proliferation in human hepatocarcinogenesis.

Deficient repair of DNA lesion O6-methylguanine in cirrhosis.

Cirrhosis is a risk factor for hepatocellular carcinoma. O6-methylguanine is a promutagenic and potentially carcinogenic DNA lesion produced by environmental alkylating agents. If it is not repaired, DNA replication can lead to a G-to-A transition mutation, which is a known mechanism of oncogene activation. We have found that the activity of the repairing methyltransferase enzyme is significantly lower in cirrhotic tissue than in non-cirrhotic diseased liver or in normal liver. This finding suggests a mechanism for cirrhosis being a risk factor for cancer of the liver: increased cellular proliferation together with persistence of O6-methylguanine might lead to malignant transformation of liver cells through mutation and oncogene activation.

Infrequent point mutations in codons 12 and 61 of ras oncogenes in human hepatocellular carcinomas.

DNA from human hepatocellular carcinomas (HCC) were analysed for the presence of mutations in codons 12 and 61 of the K-ras, H-ras and N-ras genes. The relevant ras sequences were amplified in vitro using the polymerase chain reaction and point mutations detected by selective hybridisation using mutation-specific synthetic oligonucleotides. In one of the 19 HCCs a mutation in codon 61 of the K-ras gene was detected, whilst in 3/19 HCCs a mutation was found in codon 61 of the N-ras gene. The mutations were all heterozygous A-T transversions and were found in HCCs arising in patients with underlying cirrhosis. In two of these patients where the corresponding normal tissue was available only the wild-type ras gene was detected, indicating that oncogenic activation of the ras gene was a consequence of somatic mutation. In another patient the same mutation in codon 61 of the N-ras gene was found in cirrhotic liver tissue and in all four patients the same mutation was also detected in formalin-fixed, paraffin-embedded liver biopsy HCC tissue obtained at diagnosis. These results indicate that mutational activation of the ras genes at codon 61 is an infrequent but possibly early event in the development of HCC in Britain.

c-erbB-2 oncogene expression in hepatocellular carcinoma and cholangiocarcinoma.

The c-erbB-2 proto-oncogene encodes a transmembrane protein which is homologous to the epidermal growth factor receptor. This protein can be localized immunohistochemically in formalin-fixed paraffin-embedded material using a monoclonal antibody NCL-CB11; positive membrane staining correlates with gene amplification and protein overexpression in breast cancer. Using this technique we have shown that only 2/26 (8%) of hepatocellular carcinomas, 0/10 (0%) of cholangiocarcinomas and 0/2 (0%) hepatoblastomas overexpressed c-erbB-2 as evidenced by membrane staining. Moreover c-erbB-2 mRNA was not detected in seven hepatocellular carcinomas examined by Northern blot analysis. c-erbB-2 overexpression is, therefore, unlikely to be contributing to the malignant phenotype in hepatocellular carcinoma and cholangiocarcinoma.