Faculty Bibliography

In a 65-year-old woman with cirrhosis, biphasic helical computed tomography (CT) and multiphase gadolinium-enhanced magnetic resonance (MR) imaging depicted 11 hepatic nodules that enhanced homogeneously during the hepatic arterial phase. At pathologic examination, all lesions were dysplastic nodules with unpaired hepatic arteries within. Hepatic arterial phase enhancement of cirrhotic nodules at CT and MR imaging is not diagnostic of hepatocellular carcinoma but may occur in dysplastic nodules. Results of a single biopsy yielding hepatocellular carcinoma may not be applicable to other nodules that enhance similarly

The distinction of hepatoblastoma, especially the embryonal type, from other small, round-cell tumors of childhood can sometimes be difficult. Polyclonal anticarcinoembryonic antigen (pCEA) and Hepatocyte Paraffin 1 (Hep Par 1) are immunohistochemical markers that are useful in the diagnosis of hepatocellular carcinomas. We immunohistochemically studied pCEA, monoclonal CEA (mCEA), and Hep Par 1 on 12 hepatoblastomas (3 fetal type, 2 embryonal type, and 7 mixed epithelial type). In addition, we studied the expression of Hep Par 1 on 27 other selected childhood tumors, including 1 hepatocellular carcinoma, 5 germ-cell tumors, 4 peripheral neuroectodermal tumors/Ewing's sarcomas, 3 rhabdomyosarcomas, 5 neuroblastomas, 2 rhabdoid tumors, 3 lymphomas, and 4 Wilms' tumors. All of the hepatoblastomas expressed Hep Par 1 with a characteristic granular intracytoplasmic pattern that was generally less intense in embryonal-type than in fetal-type hepatoblastomas, perhaps reflecting the degree of hepatocyte differentiation. All of the fetal-type hepatoblastomas expressed pCEA with both an intracytoplasmic and bile canalicular pattern. Embryonal type hepatoblastomas were more likely to be pCEA negative or to show focal or no canalicular pattern of expression, again possibly reflecting the degree of hepatocyte differentiation. All of the hepatoblastomas were mCEA negative. All of the nonhepatoblastomas were Hep Par 1 negative, except for the one hepatocellular carcinoma in this study, which was Hep Par 1 positive. We conclude that Hep Par 1 and pCEA are useful markers for hepatoblastomas, as they have been shown to be in hepatocellular carcinomas

We sought to determine the sensitivity and specificity of immunohistochemistry using the TORDJI-22 MoAb (BioGenex, San Ramon, Calif), which is specific for the C-100 protein of the hepatitis C virus, compared with reverse transcriptase-polymerase chain reaction (RT-PCR) of tissue for viral RNA. RT-PCR had been performed on 52 fixed tissue specimens. Immunohistochemistry was performed using prediluted antibody with the alkaline phosphatase/fast red (BioGenex) technique. Predigestion with Protease XXIV (BioGenex) and other procedures followed the manufacturer's protocols. Positive immunohistochemistry was narrowly defined as tightly clumped, perinuclear red granules in hepatocytes. Of the specimens, 28 were positive by RT-PCR. With RT-PCR as the standard of comparison, immunohistochemistry yielded a sensitivity of 70% and specificity of 84%. Positive cells, when present, were usually very rare. With stringent criteria, immunohistochemistry with the TORDJI-22 monoclonal antibody is a very specific, fairly sensitive diagnostic test for hepatitis C virus in fixed liver tissues

The blood supply of hepatocellular carcinoma (HCC) is primarily arterial. Recent studies reported differences of vascular, especially arterial, supply among low- and high-grade dysplastic nodules and HCC. We assessed arterialization using monoclonal antibody specific for smooth muscle actin as well as simultaneous changes in sinusoidal capillarization in cirrhotic nodules, dysplastic nodules, and HCC. We immunohistochemically stained 56 cirrhotic nodules, 20 low-grade dysplastic nodules, 27 high-grade dysplastic nodules, and 20 HCCs for alpha smooth muscle actin (to identify unpaired arteries (i.e., arteries not accompanied by bile ducts) and CD34 (indicating sinusoidal capillarization). Distribution and number of unpaired arteries and distribution of sinusoidal capillarization were graded semiquantitatively. Unpaired arteries were rare in cirrhotic nodules, significantly more common in dysplastic nodules of both types (p < 0.00001), and most common in HCC. Sinusoidal capillarization was least common in cirrhotic nodules, significantly more common in dysplastic nodules (p < 0.0035), and most common in HCC. No topographic relationship between unpaired arteries and sinusoidal capillarization was identified. These findings showed that (1) distributions of sinusoidal capillarization and unpaired arteries in dysplastic nodules are intermediate between those in cirrhotic nodules and HCC, supporting dysplastic nodules as premalignant lesions; (2) unpaired arteries are histologically useful for distinguishing dysplastic nodules from large cirrhotic nodules; and (3) areas of sinusoidal capillarization within dysplastic nodules are unrelated to location of arterialization

We have previously suggested that dysplastic nodules (also referred to as 'adenomatous hyperplasia' or 'macroregenerative nodules'), likely precursors of hepatocellular carcinoma (HCC), develop as an infiltrating clonal expansion, in advance of or parallel to cirrhosis. As part of this hypothesis, to explain aspects of their gross and microscopic appearance, we suggested that dysplastic nodules are resistant to the scarring process affecting the rest of the liver. We sought to test this hypothesis by examining the distribution of activated hepatic stellate cells (HSCs), the hallmark of hepatic scarring, in cirrhotic nodules, dysplastic nodules and HCC. We immunohistochemically stained 56 cirrhotic nodules, 20 low grade dysplastic nodules, 27 high grade dysplastic nodules, and 20 HCCs with monoclonal antibodies against alpha-smooth muscle actin to identify activated HSCs. Distribution and number of HSCs were graded semiquantitatively (0 to 4+). In our results, HSCs were significantly less widespread in dysplastic nodules than in cirrhotic nodules or in HCCs (both: p < 0.00001). HSCs were also more prominent in the periphery of dysplastic nodules than in the center, though still fewer in number than in cirrhotic nodules. In conclusion, the low number of activated HSCs in dysplastic nodules, compared to both cirrhotic nodules and HCC, supports our hypothesis concerning dysplastic nodule development: that they arise as clonal expansions of neoplastic hepatocytes in advance of, or parallel to, the development of cirrhosis