The Sialyl Lewis X expression in hepatocarcinogenesis: potential predictor for the emergence of hepatocellular carcinoma

BACKGROUND/AIMS: Sialyl Lewis X (sLeX), one of the cancer-associated glycoproteins, has been reported to be expressed in both liver tissue from various types of liver disease and hepatocellular carcinoma. In order to clarify sLeX expression during the early stage of hepatocarcinogenesis, we examined sLeX expressions in either liver tissue specimens without any nodular lesions, dysplastic nodules or hepatocellular carcinomas. METHODOLOGY: Immunohistochemical observations were performed using a monoclonal antibody for sLeX. As for the liver tissue specimens, 8 livers without any chronic liver disease and 42 diseased livers were examined, while for the nodular lesions, 5 dysplastic nodules (borderline lesions) and 47 hepatocellular carcinomas were examined in this study. RESULTS: sLeX was not expressed in all 8 normal livers. sLeX was expressed membraneously in 8 of 15 (53%) chronic hepatitic liver tissue specimens, in 8 of 9 (89%) precirrhotic liver tissue specimens and in 16 of 18 (89%) cirrhotic liver tissue specimens. The incidence of sLeX expression on hepatocytes in both pre-cirrhotic and cirrhotic liver tissue was higher than that in chronic hepatitic liver tissue (P < 0.05). The sLeX expression in liver tissue was positive in all 14 liver tissue specimens containing multiple hepatocellular carcinomas, in which at least one of the nodules was a well-differentiated hepatocellular carcinoma regarded as multicentric hepatocellular carcinomas. In 18 of 28 (64%) liver tissue specimens without multicentric hepatocellular carcinomas, sLeX was positive and the difference was statistically significant (P < 0.05). In nodular lesions, sLeX was negative in 5 dysplastic nodules (borderline lesions). In hepatocellular carcinoma, 14 of 47 (30%) hepatocellular carcinoma nodules showed a positive expression. Six of 14 (43%) well-differentiated hepatocellular carcinomas were positive on the cell membrane. Four of 23 (17%) moderately differentiated hepatocellular carcinomas were positive on the cell membrane, while one of 23 (4%) moderately differentiated hepatocellular carcinoma was positive in the cytoplasm. In addition, 3 of 10 (30%) poorly differentiated hepatocellular carcinomas were positive in the cytoplasm. CONCLUSIONS: These results suggested that the sLeX-positive liver tissue specimens possessed a high degree of carcinogenicity and therefore sLeX expression in the diseased liver might be a good predictor for hepatocellular carcinoma emergence. At the same time, the suppression of sLeX occurred at a very early stage of hepatocarcinogenesis. In addition, the phenotype of sLeX was also considered to change during the progression of hepatocarcinogenesis.     

Incidental solitary hepatocellular carcinomas smaller than 1 cm in size found at autopsy: a morphologic study

The morphologic features and growth pattern of single hepatocellular carcinomas less than or equal to 1 cm in size, found incidentally at autopsy, were studied in nine cases. In all but one case, the hepatic parenchyma showed advanced cirrhosis. In three cases, the hepatocellular carcinomas were localized within a regenerative nodule as a form of "nodule within nodule." The carcinoma was rimmed by nonneoplastic hepatic tissue. A fourth carcinoma consisted of an expansile hepatocellular carcinoma nodule enclosed by a fibrous band of cirrhotic stroma. The remaining five cases consisted of hepatocellular carcinomas which infiltrated the surrounding regenerative nodules or hepatic lobules. These observations suggested that hepatocellular carcinomas arise within regenerative nodules, some of which still retain residual nonneoplastic tissue around the tumor. Others invaded the adjacent liver tissue. The grossly visible fibrous capsule, often seen in more advanced hepatocellular carcinomas, was absent in all cases. All of the hepatocellular carcinomas were well-differentiated. Four cases showed a trabecular pattern with slight sinusoidal dilatation, 3 showed a scirrhous pattern and 2 showed a compact pattern. Their histologic features included marked bile production, Mallory body formation by clusters of tumor cells, resistance to hemosiderin deposition in a markedly siderotic background and loss or decrease of reticulin fibers. These features were hallmarks of small hepatocellular carcinomas. Pathologists should study cirrhotic livers carefully so as not to miss small carcinomas. Clinicians should be aware that even small liver nodules may be hepatocellular carcinomas.     

Ultrasonic Doppler studies of hepatocellular carcinoma and comparison with other hepatic focal lesions

One hundred fifty-four liver lesions, including 63 hepatocellular carcinomas, were studied to determine the value of duplex ultrasound in the diagnosis of small hepatocellular carcinomas. Arterial Doppler signals were obtained either within the body of the tumor, at its periphery, or in both locations, from 28 to 37 hepatocellular carcinomas less than or equal to 3 cm in diameter and from all 26 hepatocellular carcinomas with a diameter greater than 3 cm. Arterial Doppler signals were obtained at the periphery of 5 of 7 cholangiocarcinomas, 4 of 11 liver metastatic tumors, and 5 of 23 hemangiomas. No such signals were obtained from 29 regenerative nodules, 10 hepatic pseudotumors, and 11 liver cysts. The mean peak systolic frequency seen in hepatocellular carcinoma (1.2 kHz) was significantly greater than in cholangiocarcinoma (0.6 kHz), metastatic tumors (0.5 kHz), or hemangiomas (0.3 kHz). A peak systolic frequency of greater than 3 kHz was found in 6 of 8 hepatocellular carcinomas greater than or equal to 4 cm in diameter with angiographically proven arterioportal shunting, whereas the value in other hepatocellular carcinomas or other hepatic focal lesions was less than 2.6 kHz. This study showed that the peak systolic shift was related to the degree of arterioportal shunting. Because shunting is either minor or nonexistent in small hepatocellular carcinomas, the value of duplex Doppler ultrasound in the diagnosis of these lesions appears to be limited.     

A histopathological study on combined hepatocellular and cholangiocarcinoma: cholangiocarcinoma component is originated from hepatocellular carcinoma

BACKGROUND/AIMS: Combined hepatocellular and cholangiocarcinoma of the liver is relatively infrequent, and its pathogenesis remains obscure. The aim of this study is to investigate its clinical and pathological features and proliferative activity. METHODOLOGY: In this study, we investigated the histopathological features, Ki-67 labeling index, and p53 immunohistochemistry of 18 surgically resected cases of combined hepatocellular and cholangiocarcinoma among 1102 consecutive cases of surgically resected primary liver cancers. All tumors were compatible with the WHO definition of this tumor. Microscopically, we classified the cases into the following three categories according to the arrangement of the hepatocellular carcinoma and cholangiocarcinoma components; (1) Type I in which hepatocellular carcinoma and cholangiocarcinoma formed nodules that could easily be distinguished from each other, (2) Type II in which the both components were finely mixed, so that the two components were almost indistinguishable, and (3) Type III in which the tumors had lobular structures with hepatocellular carcinomas existing centrally and cholangiocarcinomas existing peripherally. RESULTS: Microscopically, the tumors were classified into type I 7 tumors, type II 5 tumors, and type III 6 tumors. In one case of type I, well differentiated hepatocellular carcinoma demonstrated cholangiocarcinoma in "nodules-in-nodules" fashion. The average of Ki-67 labeling index of hepatocellular carcinoma component of combined hepatocellular and cholangiocarcinoma was 4.4 +/- 3.4% and the index of cholangiocarcinoma component was 11.0 +/- 8.5%, which is significantly higher than that of the hepatocellular carcinoma component. On p53 immunohistochemistry, 5 of 18 cases (29.4%) were positive. In one case, the cholangiocarcinoma component was positive for p53, but the hepatocellular carcinoma component was negative. In the other 4 cases, both the hepatocellular carcinoma and cholangiocarcinoma components were positive. CONCLUSIONS: Microscopically, type III seems to be a feature of metaplasia or proliferation of bipotential progenitor cells. Metaplasia of hepatocellular carcinoma to intrahepatic cholangiocarcinoma is assumed to be one of the pathogenic pathways of combined hepatocellular and cholangiocarcinoma.     

Small hepatocellular carcinoma with minute satellite nodules

BACKGROUND/AIMS: To investigate the clinicopathologic characteristics of small hepatocellular carcinoma with minute satellite nodules. METHODOLOGY: We investigated the clinicopathologic characteristics of 131 solitary small (< or = 2.0 cm in diameter) hepatocellular carcinomas including 105 hepatocellular carcinomas without minute satellite nodules and 17 hepatocellular carcinomas with minute satellite nodules smaller than 5 mm, and also discuss the clinical significance. RESULTS: None of the clinical backgrounds of the patients and pathologic features of the main tumor, except for the average of preoperative serum alpha-fetoprotein, were significantly different between the two groups. Firstly, minute satellite nodules demonstrated that the maximum diameter of all minute satellite nodules was 1.5-4.0 mm, secondly, the moderately to poorly differentiated hepatocellular carcinomas had 4 or more minute satellite nodules within 1 cm from the main tumor, while well differentiated hepatocellular carcinomas may have 1 or 2 minute satellite nodules 6 cm or more away, and thirdly, 4 or more minute satellite nodules may present within 1 cm in intrahepatic metastasis cases, while 1 or 2 minute satellite nodules may be present 6 cm or more away from the main tumor in multicentric occurrence cases. CONCLUSIONS: At least 13% of solitary small hepatocellular carcinomas had preoperatively undetectable minute satellite nodules. In case of moderately to poorly differentiated hepatocellular carcinomas, hepatic resection as well as percutaneous ethanol injection should be performed including the surrounding liver tissue at least 1.0 cm from the main nodule. On the other hand, in well-differentiated hepatocellular carcinomas, which may indicate multicentric occurrence, closer observation and careful follow-up after therapy are recommended.     

Incidence and diagnostic features of macroregenerative nodules vs. small hepatocellular carcinoma in cirrhotic livers.

In Japan, the presence of a large regenerative nodule within a cirrhotic liver, referred to as a macroregenerative nodule or adenomatous hyperplasia, is thought to play a role in the pathogenesis of hepatocellular carcinoma. These lesions, however, have received little attention outside of Japan. We examined 110 sequentially explanted cirrhotic livers for the presence of such nodules. By gross examination, 19 livers (17.3%) had 40 nodules (10 livers had more than one nodule) between 0.8 and 3.5 cm in diameter. By histological examination, 28 of these were macroregenerative nodules and 12 were hepatocellular carcinomas. Three of these hepatocellular carcinomas, however, appeared to have arisen in association with a macroregenerative nodule. We found that the architectural features of thickened cell plates, formation of trabeculae and loss of reticulin were usually very helpful in differentiating benign macroregenerative nodules from hepatocellular carcinoma. The incidence of macroregenerative nodules in our series was similar to that seen in the Japanese studies, and although we feel that they may play a role in the pathogenesis of carcinoma, we do not believe their presence is necessary for the development of hepatocellular carcinoma.     

Loss of fragile histidine triad protein in human hepatocellular carcinoma

AIM: To investigate the expression of fragile histidine triad (FHIT) gene protein, Fhit, which is recently thought to be a candidate tumor suppressor. Abnormal expression of fragile histidine triad has been found in a variety of human cancers,but little is known about its expression in human hepatocellular carcinogenesis and evolution.METHODS: Sections of 83 primary human hepatocellular carcionoma with corresponding para-neoplastic liver tissue and 10 normal liver tissue were evaluated immunohistochemically for Fhit protein expression.RESULTS: All normal liver tissue and para-neoplastic liver tissue showed a strong expression of Fhit, whereas 50 of 83(65.0 %) carcinomas showed a marked loss or absence of Fhit expression. The differences of Fhit expression between carcinoma and normal or para-neoplastic liver tissue werehighly significant (P=0.000). The proportion of carcinomas with reduced Fhit expression showed an increasing trend (a) with decreasing differentiation or higher histological grade (P=0.219); (b) in tumors with higher clinical stage Ⅲ and ⅣV (91.3 %, P=0.000), compared with tumors with lower stage Ⅰ and Ⅱ (27.6 %); and (c) in cancers with bigger tumor size (＞50 mm) (75.0 %, P=0.017), compared withsmaller tumor size (≤ 50 mm). CONCLUSION: FHIT inactivation seems to be both an earlyand a later event, associated with carcinogenesis andprogression to more aggressive hepatocellular carcinomas.Thus, evaluation of Fhit expression by immunohistochemistryin hepatocellular carcinoma may provide important diagnosticand prognostic information in clinical application.     

Resistance to iron accumulation and presence of hepatitis B surface antigen in preneoplastic and neoplastic lesions in human hemochromatotic livers

Two human cases of hepatocellular carcinoma combined with primary hemochromatosis and liver cirrhosis were studied with special reference to liver lesions displaying resistance to iron accumulation, and presence of hepatitis B surface antigen. Hepatocellular carcinomas, as well as small islands of hepatocytes within regenerative nodules, were free of the iron accumulation which otherwise occurred throughout the remainder of the hemochromatotic liver parenchyma. A positive reaction for hepatitis B surface antigen using the orcein staining method occurred randomly in iron-containing hepatocytes and in clusters of iron-containing hepatocytes cirrhotic nodules, but completely iron-free cells of foci and carcinomas were negative for orcein. Therefore, these iron-free foci are suggested to be precursors to the carcinoma.     

Occurrence of oval-type cells in hepatitis B virus-associated human hepatocarcinogenesis.

Proliferation of a new population of epithelial cells with distinct structure, as well as cytokeratin and alpha-fetoprotein expression, was observed in nonneoplastic liver tissues from 14 cases (13 hepatitis B virus-positive) of human hepatocellular carcinoma. These cells were characterized by oval nuclei; scant, pale cytoplasm; small cell size; and cross-reaction with a monoclonal antibody against rat oval cells. These putative human oval cells were strongly positive for cytokeratin 19 and displayed considerable heterogeneity in alpha-fetoprotein and albumin expression. The oval cells were most prominent in actively regenerating nodules and in liver tissue surrounding the cancer. Oval cells and transitional types of cells appear to be the principal producers of alpha-fetoprotein in the regenerating liver. Cancer cells positive for cytokeratins 8, 18 and 19 were observed in half the hepatocellular carcinomas studied. The data suggest that a new cell population structurally similar to oval cells seen in early stages of chemical hepatocarcinogenesis in rats is consistently present in regenerating liver lesions associated with human hepatocellular carcinoma. Furthermore, it is possible that the proliferation of these oval-type cells may partly account for the elevation of serum alpha-fetoprotein frequently seen in precancerous stages of hepatitis B virus-associated human hepatocellular carcinoma.     

Clonal expansion in evolution of chronic hepatitis to hepatocellular carcinoma as seen at an X-chromosome locus

Clonal analysis has shown that hepatocellular carcinoma arises from a single cell. However, the clonality of precancerous lesions and adjacent nonneoplastic tissues is not clear. We analyzed a human androgen receptor locus to elucidate the clonal state of liver tissues including post-hepatitic lesions associated with hepatocarcinogenesis. The analysis was based on a restriction fragment length polymorphism involving an androgen receptor locus on the X chromosome, taking advantage of physiologic random inactivation by methylation of 1 of 2 X chromosomes in females during embryogenesis. Clonality was assessed in 79 randomly located tissue samples microdissected from noncirrhotic liver, including a total of 40 morphologically normal sites in 4 normal livers and 39 sites from a single HCV-infected liver. In addition, 51 regenerative nodules, 4 areas of adenomatous hyperplasia, and 18 hepatocellular carcinomas were sampled. All samples were obtained from livers involved by various neoplasms. Eight of forty samples (20.0%) from the four normal livers and 20 of the 39 samples (51.3%) from the single HCV-infected liver showed a monoclonal pattern. Moreover, 30 of 51 regenerative nodules (58.9%) showed a monoclonal pattern. No histologic differences were evident between mono- and polyclonal nodules. On the other hand, the 18 carcinomas and 4 areas of adenomatous hyperplasia all were monoclonal. Mean calculated monoclonal areas of normal liver and liver with chronic hepatitis were 1.1 and 3.3 mm(2). Our results suggest that areas representing a single clone of hepatocytes are present in normal liver, and these progressively expand as changes advance from chronic hepatitis to hepatocellular carcinoma.     

Imaging diagnosis of hepatocellular carcinoma]

In patients with chronic liver disease, hepatocellular carcinomas are developed from regenerative nodule via dysplastic nodule and early hepatocellular carcinoma to advanced hepatocellular carcinoma during multistep hepatocarcinogenesis. In this article, imaging findings of various imaging modalities are described pertaining to the above mentioned hepatocellular nodules occurring in the cirrhotic liver, correlating with pathologic findings.     

Angiographically undetected small hepatocellular carcinoma: clinicopathological characteristics, follow-up and treatment

We studied 36 cases (38 nodules) of small hepatocellular carcinoma with special attention directed to detectability using angiography and histology. Among the 38 nodules, 31 nodules (81.6%) were evident. The remaining seven (18.4%) were not evident angiographically but were detected using ultrasonography and/or computed tomography, prior to angiography. An elevated level of serum alpha-fetoprotein also suggested a diagnosis of hepatocellular carcinoma. When comparisons were made between angiographically detected and undetected nodules, there was no statistically significant difference in parameters, including sex of the patient, tumor size, location of the tumor, positive rate for hepatitis B surface antigen, history of liver disease and values of alpha-fetoprotein. The small hepatocellular carcinomas not detected angiographically and evidenced only histologically had the following characteristics: (i) no fibrous capsule surrounding the nodules; (ii) well-differentiated tumors; (iii) replacing growth pattern of cancer cells, and (iv) remains of portal tracts within the tumor. The replacing growth pattern and the presence of portal tracts may correlate with the low detectability in angiography. When a definite diagnosis of angiographically undetectable tumors cannot be made using close surveillance with ultrasonography or alpha-fetoprotein and/or needle biopsy under ultrasonic guide, surgical intervention should be considered, particularly for patients at high risk for hepatocellular carcinoma.     

5'-Nucleotide phosphodiesterase isozyme-V in hepatocellular carcinoma and other liver diseases.

The clinical usefulness of serum 5'-nucleotide phosphodiesterase isozyme-V (5'-NPD-V) assay as a serological marker for hepatocellular carcinoma was evaluated. Serum levels of 5'-NPD-V were measured by polyacrylamide gel electrophoresis in 536 Japanese patients with various diseases, including 120 patients with hepatocellular carcinoma. Icteric serum was not an indicator for the measurement of this isozyme, because jaundice gave a non-specific false-positive reaction. In 99 cases of hepatocellular carcinoma without jaundice, 73 (74%) had positive 5'-NPD-V and 24 (24%) showed levels greater than (+ +). The diagnostic value of this isozyme for hepatocellular carcinoma was relatively high, especially in patients with low or negative AFP levels. Diagnostic application for serum 5'-NPD-V assay to small liver tumors was limited. 5'-NPD-V showed false-positive results even in certain cases of benign liver diseases such as chronic hepatitis and liver cirrhosis, but cases with positivities stronger than (+ +) were few. Moreover, the test might be useful for the prediction of liver metastasis in cancer patients, since positive rates were significantly higher in cases with liver metastasis than in those in the non-liver metastasis group.     

The value of ultrasonography for hepatic surgery.

Ultrasonography plays an important role in the early detection of hepatocellular carcinoma. Ultrasonography detected 53% of 287 patients with small (less than or equal to 5 cm) hepatocellular carcinomas. Among 486 patients with hepatocellular carcinomas, sensitivities of intra-operative ultrasonography in detecting 451 small primary hepatocellular carcinoma nodules, intrahepatic metastasis in 330 patients with small hepatocellular carcinoma and 63 tumor thrombi were 98, 48 and 67%, respectively. The sensitivity of intra-operative ultrasonography in detecting small primary tumors was 10% better than ultrasonography, computed tomography, and angiography. The sensitivity of intra-operative ultrasonography in intrahepatic metastasis and tumor thrombus was two to three times better than pre-operative examination. Intra-operative ultrasonography was useful in detecting nonpalpable tumors and in guiding the transection of the liver, biopsy, and cryosurgery. Moreover, intra-operative ultrasonography made possible new hepatectomy procedures: systematic subsegmentectomy and hepatectomies which preserve the inferior right hepatic vein. Systematic subsegmentectomy guided by intra-operative ultrasonography resulted in better survival rates than the limited resection in patients with small hepatocellular carcinoma two years after hepatectomy; by the sixth year, this difference was significant (p less than 0.05). Ultrasonography and intra-operative ultrasonography are indispensable in the early detection, accurate diagnosis, operative guidance and postoperative care of hepatocellular carcinoma.     

Phase II study of systemic gemcitabine chemotherapy for advanced unresectable hepatobiliary carcinomas.

BACKGROUND/AIMS: Patients with advanced unresectable hepatobiliary carcinomas have a dismal prognosis. The efficacy of systemic chemotherapy in these patients is negligible and often, in particular in patients with hepatocellular carcinomas, the toxicity of chemotherapy outweighs the potential palliative effect of antineoplastic agents. Gemcitabine is a new anticancer agent with a mild toxicity profile, which has demonstrated antineoplastic activity in many solid tumors. Therefore we investigated the effect of gemcitabine in patients with advanced nonresectable hepatocellular and cholangiocellular carcinomas in a phase II study. METHODOLOGY: Twenty-three patients with cholangiocellular carcinoma and 20 patients with hepatocellular carcinoma were enrolled into the study. Eighteen of the 20 patients with hepatocellular carcinomas had liver cirrhosis. Gemcitabine was administered once weekly over 30 min for 3 consecutive weeks out of every 4 weeks. Patients with cholangiocellular carcinomas received gemcitabine also in the forth week of the first cycle with no rest to the following cycle. Disease status was assessed every 4 weeks. RESULTS: Overall the regimen was well tolerated. The median number of gemcitabine administration was 15 (range, 3-37) in the group of patients with cholangiocellular carcinomas and 7.6 (range, 3-21) in the group of patients with hepatocellular carcinomas. In the group of patients with hepatocellular carcinomas thrombocytopenia was the most frequent side effect (30% grade 3/4). Among the patients with cholangiocellular carcinomas nausea and neutropenia were the most commonly observed side effects. The overall response rate of hepatocellular carcinomas was only 5% and chemotherapy generally did not improve the tumor symptoms of the patients in this group. In contrast, in the group of cholangiocellular carcinomas, seven patients achieved a partial response (overall response rate 30%). Eleven patients with cholangiocellular carcinomas revealed tumor symptoms before the onset of gemcitabine treatment. Seven of these patients developed a treatment related clinical benefit as defined as a relief of tumor symptoms or gain of weight. CONCLUSIONS: Our results indicate that the treatment of cholangiocarcinomas with gemcitabine is effective and should be further evaluated in phase III studies. In contrast, palliative chemotherapy with gemcitabine cannot be recommended in patients with hepatocellular carcinoma and liver cirrhosis.     

Liver tumors and possible preneoplastic lesions,induced by a food-derived heterocyclic amine in cynomolgus monkeys; a study of histology and cytokeratin expression

Abstract: A food-derived mutagenic heterocyclic aromatic amine, 2-amino-3-methylimidazo[4,5-f]quinoline (IQ), is a potent hepatocarcinogen in cynomolgus monkeys. In an ongoing carcinogenesis study, 34 out of 40 monkeys dosed with IQ have developed malignant liver tumors. The histology and cytokeratin expression was examined in a total of 94 tumors and non-neoplastic lesions obtained from 34 cases. The majority of the tumors were classified as hepatocellular carcinoma. In some cases, a striking difference in the histological features between individual tumor nodules was suggestive of a multicentric origin. Intrahepatic vascular invasion was seen in 14 (41.2%) and metastases in 6 (17.6%) of the hepatocellular carcinoma cases. There was no evidence of regenerative hyperplasia or fibrosis in the parenchyma of the tumor-bearing livers. Clear-cell foci composed of glycogen-rich hepatocytes were the only macroscopic lesions detected prior to gross tumor development. Other liver lesions included dysplastic hepatocyte foci and areas of proliferating bile ductular like (oval) cells, located around the periportal areas and along the portal tracts. Expression of bile duct type cytokeratin 7 was observed in a few of the oval cells and non-malignant hepatocytes, as well as in some of the hepatocellular carcinoma nodules. This aberrant cytokeratin expression raises questions concerning the histogenesis of the IQ-induced hepatocellular carcinoma.