Background: The cytopathologic distinction between hepatocellular carcinoma (HCC) and metastatic carcinoma (MC) in the liver can be problematic, especially in patients with poorly differentiated HCC, in whom a trabecular pattern, bile production, and Mallory bodies may not be apparent on small fine-needle aspiration (FNA) samples. HepPar1 (OCH1E5) is a monoclonal antibody specifically developed to react with hepatocytes. It rarely reacts with bile duct and nonparenchymal liver cells.
Methods: FNA samples (cell blocks) from 75 liver tumors were selected. These included 50 moderate to poorly differentiated HCC cases, 5 cholangiocarcinoma (CC) cases, and 20 MC cases (4 from the breast, 4 from the stomach, 4 from the pancreas, and 8 from the colon). Immunohistochemical staining for HepPar1 was performed to differentiate HCC from MC.
Results: The HepPar1 antibody was positive in 50 of 50 HCC cases (100%). The positivity was cytoplasmic, diffuse, and granular. All 5 cases of CC were found to be negative (0%). Although focal positivity within tumor cells was noted in one case, cytologically these were entrapped normal hepatocytes between the tumor cells. In addition, 3 of 20 MC cases (15%) also were positive for HepPar1. All three cases originated from gastric primary tumors and exhibited diffuse, granular cytoplasmic staining.
Conclusions: The results of the current study demonstrate that HepPar1 is an effective marker with which to differentiate between HCC and CC and/or MC. HepPar1 was found to demonstrate 100% positivity in HCC cases, compared with 0% and 15% positivity, respectively, in CC and MC cases. In addition, HepPar1 is extremely helpful in limited tissue samples from FNA. Although 15% of the MC cases in the current study were found to be positive, with the help of clinical correlation and other immunohistochemical stains a definite diagnosis could be rendered. Potential pitfalls include residual benign hepatocyte staining within a non-HCC malignancy, as was observed in one of the CC cases in the current study.