Infantile Liver Giant Cells: Immunohistological Study of Their Proliferative State and Possible Mechanisms of Formation

Abstract

The mechanism of liver giant cell formation is not clarified. Some authors consider the giant cells regenerative, others, degenerative. Paraffin sections of 10 archival cases of idiopathic neonatal hepatitis (INH), 8 of extrahepatic biliary atresia (EHBA), and 5 normal liver samples were immunostained with two well-characterized cell proliferation markers: anti-PCNA monoclonal antibody (MAb) (clone PC-10) and MAb MIB-1, which detects Ki-67, a nuclear proliferation-related antigen. In addition, polyclonal antibody to carcinoembryonic antigen (CEA) was used to identify remnants of canalicular, therefore hepatocytic, membranes in giant cells. Quantitative analysis of immunostaining was done by estimating PCNA and Ki-67 indices separately in giant cells and in nongiant hepatocytes. In normal samples, mean PCNA and Ki-67 indices were 1.22% and 0.74%, respectively. In the cases of INH and EHBA, only a small minority of giant cells showed PCNA or Ki-67 staining limited to occasional peripherally located nuclei. PCNA and Ki-67 indices were significantly higher in the non–giant cell compartment. CEA staining was seen only in rare giant cells as centrally located canalicular remnants bordered by polarized nuclei, suggesting that they had been formed from rosettes through dissolution of cell membranes. Other giant cells shared CEA-labeled canalicular membranes with mononuclear hepatocytes in rosettes. These findings indicate that the giant cells in INH and EHBA are not regenerative cells, they are not formed by amitotic division of nuclei in syncytia, and that fusion of rosette-forming hepatocytes is a possible mechanism of their formation.