Giant cell tumor of the pancreas is a very rare neoplasm and appears as two different phenotypes: undifferentiated carcinoma and osteoclast-like giant cell tumor (OGTP) (1-3). OGTP, mimicking giant cell tumor of bone should be distinguished from undifferentiated carcinoma. In contrast to OGTP, undifferentiated carcinoma exhibit more pleomorphic and anaplastic features. OGTP is the more rare type and has a better prognosis than undifferentiated carcinoma (1-3). An implication of these differences in histological and clinical characteristics is that these tumors have a different histogenesis. To date, generally, undifferentiated carcinoma is believed to have an epithelial origin without debating. However, the histogenesis of OGTP remains controversial either epithelial or mesenchymal derivation in the literature (4-16).

Here, we report a case of OGTP, performed the histopathological, immunohistochemical, ultrastructural, and molecular biological studies and review the literature pertaining to this condition.

A 63-yr-old woman was admitted to Chonnam National University Hospital with a three month history of epigastric pain. Physical examination showed a smooth palpable mass in the epigastrium. The laboratory examinations were within normal range. An abdominal computerized tomography revealed the presence of a large cystic mass, arising from the tail of pancreas (Fig. 1A). A distal pancreatectomy with splenectomy was performed. The resected tumor was a 8×7×6-cm hemorrhagic and cystic mass in the tail of the pancreas, mimicking pancreatic pseudocyst (Fig. 1B). The histology confirmed OGTP with ductal adenocarcinoma. She received third courses of chemotherapy with gemcitabine. She has been well, without evidence of tumor recurrence 6 months after surgery.

OGTP is very rare and comprises less than 1% of nonendocrine pancreatic tumor (1-3). To date, only two cases have been reported in Korea (18, 19). In the review of English literature (20-22), OGTP tends to present in the 7th decade of life, although younger patients have been reported. There was no sex preference. The main symptoms and signs were abdominal pain, palpable mass, weight loss and jaundice. OGTP commonly exhibited as large cystic neoplasms with hemorrhage and necrosis of variable extent. There was predilection to involve head and body portion of the pancreas. The prognosis of OGTP is better than that of pancreatic ductal adenocarcinoma and undifferentiated carcinoma, suggesting that OGTP grows more slowly and is less likely to metastasize than pancreatic ductal adenocarcinoma and undifferentiated carcinoma. However, the outcome of OGTP is extremely variable in the literature. Although it is difficult to decide treatment modality because of its rarity, the choice of treatment of OGTP may be surgical resection if possible. Also, in previous reports, experience of adjuvant treatments including chemotherapy or radiotherapy was not described, so no detailed information was provided. However, because the outcome of OGTP is extremely variable, it may be reasonable to consider adjuvant treatment for prevention of tumor recurrence. In our case, we did attempt chemotherapy with gemcitabine, according to treatment protocol established for pancreatic ductal adenocarcinoma. We think that this attempt is important and necessary to detail the treatment modalities and clinical outcome.

Although the histiogenesis of OGTP has been previously investigated by immunohistochemical, ultrastructural, and molecular biologic studies, the origin of these cells is still controversial.

In our case, osteoclast-like giant cells and mononuclear cells were positive for histiocytic markers, CD68 and lysozyme in contrast to negative for epithelial markers, CK and EMA. In the ultrastructure of multinucleated, osteoclast-like giant cells and mononuclear cells, we did not find either microvilli or desmosomes for the clue of epithelial differentiation.

Tumor progression is regulated by many biological processes including cell proliferation and genetic alterations of tumor suppressor gene and oncogene. PCNA reflects cell proliferation and mutation of the p53 tumor suppressor gene represents one of the common genetic alterations in pancreatic ductal carcinoma. In our case, ductal carcinoma cells expressed PCNA while no nuclei were stained in osteoclast-like giant cells and mononuclear cells. p53 immunoreactivity showed only weak intranuclear staining in some atypical mononuclear cells, but did not show in ductal carcinoma cells and osteoclast-like giant cells.

K-ras gene mutations are a hallmark of pancreatic ductal carcinoma and appear to be present in 80% to 90% of tumors. Evidence of K-ras gene mutations indicates that its origin is an epithelial nature. But in our case, none of the microdissected osteoclast-like giant cell, mononuclear cell, and ductal carcinoma cell contained K-ras gene mutations.

These results may support that both osteoclast-like giant cells and mononuclear cells in OGTP originate from mesenchymal rather than epithelial cell. However, although most reports have debated the histiogenesis of OGTP from immunohistochemical, ultrastructural, and molecular biologic aspects, it is still obscure whether the histiogenesis of OGTP is epithelial or mesenchymal origin (4-16). There are several possible explanations for these discrepancies. First, OGTP is too rare to discern differences in the origin of these cells. Second, immunohistochemical staining methods and primary antibodies used are different in different cases (4). Third, OGTP can express different phenotypes with variable stages and degrees of differentiation during tumor progression (4, 23). Fourth, circulating precursor cell may be stimulated by tumor derived factors such as cytokines, to proliferate and differentiate into different cell types (4, 16, 20, 21).

Clinical characteristics and histiogenesis of OGTP are still now variable and controversial in the literature. Therefore, further studies based on a large populations are necessary to clarify its histiogenesis.