Ovarian-Type Stroma in Hepatobiliary Cystadenomas and Pancreatic Mucinous Cystic Neoplasms An Immunohistochemical Study

Transcription

1 Anatomic Pathology / OVARIAN-TYPE STROMA IN LIVER AND PANCREATIC LESIONS Ovarian-Type Stroma in Hepatobiliary Cystadenomas and Pancreatic Mucinous Cystic Neoplasms An Immunohistochemical Study Maggie M. Lam, MD, Paul E. Swanson, MD, Melissa P. Upton, MD, and Matthew M. Yeh, MD, PhD Key Words: Hepatobiliary cystadenoma; Pancreatic mucinous cystic neoplasm; α-inhibin; c-met; Hepatocyte growth factor DOI: /U2BBP4EMBAHCM6E6 Abstract We compared immunohistochemical expression of hepatocyte growth factor (HGF), c-met, α-inhibin, estrogen receptor (ER), and progesterone receptor (PR) in 10 pancreatic mucinous cystic neoplasms (PMCNs), 8 hepatobiliary cystadenomas (HBCs), and 6 simple liver cysts (SLCs). All HBCs and PMCNs were in women (mean ages, 45.7 and 54.9 years, respectively; P >.05). All HBCs had abundant stroma; PMCN stromal density was more variable (score, 3.0 vs 1.8; P <.05). Stroma in HBCs had greater ER and PR staining (both P <.05) and more consistent, focal, strong reactivity for α-inhibin than PMCNs (P <.05). SLCs were found in men and women, lacked ovarian-type stroma, and were negative for ER, PR, and α-inhibin. HGF expression was not significantly different in the epithelium of HBCs, PMCNs, and SLCs; c-met expression in the epithelium of HBCs and PMCNs was significantly stronger than in SLCs (P <.05). Differences in stromal density and ER, PR, and α- inhibin immunoreactivity in HBCs and PMCNs suggest that different hormonal environments in the liver and pancreas contribute to stromal variation. Up-regulation of c-met in HBCs and PMCNs suggests c-met activation in the pathogenesis of pancreaticobiliary cystic neoplasms. Primary cystic lesions of the pancreas and liver include pancreatic mucinous cystic neoplasms (PMCNs) and hepatobiliary cystadenomas (HBCs), respectively. Although primary to different organs, these lesions are strikingly similar, and each resembles ovarian mucinous cystic neoplasms. HBCs are cystic, multilocular lesions lined by benign, simple columnar or cuboidal glandular epithelium. They are distinguished from simple liver cysts and the recently described cystic variants of intraductal papillary neoplasms of the bile duct by the presence of subepithelial stroma that consists of closely packed spindled cells with an appearance reminiscent of ovarian stroma. 1 HBCs occur primarily in women, suggesting a sex-specific factor in their pathogenesis. 2,3 Similarly, PMCNs, initially elucidated by Compagno and Oertel, 4 are chiefly found in women and have histologic features that are similar to HBCs, again emphasizing an ovarianlike stroma. The sex specificity of PMCNs, as well as their anatomic location (predilection for the tail of the pancreas and no communication with the pancreatic ductal system) and ovarian-like stroma, help to distinguish them from cystic intraductal neoplasms of the pancreas, including intraductal papillary mucinous tumors 5 and pancreatic intraductal neoplasia. The ovarian-like stroma of HBCs and PMCNs expresses estrogen receptor (ER) and progesterone receptor (PR). 6-8 α- Inhibin, a gonadal protein that has a more limited expression in the sex cord stromal tissue, is also expressed in PMCNs and sex cord stromal tumors. 5,9-13 Its expression has not yet been characterized in HBCs. Hepatocyte growth factor (HGF) and its tyrosine kinase receptor (c-met) are important signaling pathways for the proliferation of intrahepatic biliary epithelial cells, 14 hepatocellular carcinoma, 15 and other non liver-specific tissues, such as Am J Clin Pathol 2008;129: DOI: /U2BBP4EMBAHCM6E6 211

2 Lam et al / OVARIAN-TYPE STROMA IN LIVER AND PANCREATIC LESIONS ovarian surface epithelial neoplasms. 16 Based on these observations, it is likely that HGF and c-met contribute to the pathogenesis of HBCs and PMCNs. The aims of the present study were to reexamine the overlapping features and the clinicopathologic differences between HBCs and PMCNs and to study the immunohistochemically defined expression of HGF, c-met, inhibin, ER, and PR in PMCNs and HBCs as a means to further elucidate the morphogenesis and pathogenesis of these uncommon cystic lesions. Materials and Methods Patients and Samples The study protocol was reviewed and approved by the Human Subject Division, University of Washington, Seattle. The files of the Department of Pathology, University of Washington, between January 1982 and December 2004 were searched for cases diagnosed as HBC or PMCN. Archived cases of 8 HBCs, 10 PMCNs, and, for comparative purposes, 6 simple liver cysts were identified. The clinical charts, pathology reports, and slides were reviewed. Histologic Examination H&E-stained slides were reviewed. The presence of mucin-producing epithelial cells lining an ovarian-type stroma was confirmed in each case of cystic neoplasms. The stromal cell density was semiquantitatively graded from 0 to 3 (0, no stroma; 1, scant stroma; 2, intermediate stroma; and 3, abundant stroma). Features typical of epithelial dysplasia (ie, enlargement of cells, multilayering, nuclear hyperchromasia, and loss of nuclear polarity) and overt malignancy (ie, invasion of the underlying stroma, back-to-back or complex glands, nuclear pleomorphism, and increased mitotic activity) were evaluated in each case. 2 Immunohistochemical Analysis Formalin-fixed, deparaffinized 5-µm sections were incubated with the following panel of monoclonal and/or polyclonal antibodies: ER (clone 1D5, dilution 1:250; Immunotech/Coulter, Miami, FL), PR (PR88, dilution 1:250; BioGenex Laboratories, San Ramon, CA), inhibin (Inhibin Alpha; dilution 1:10; Serotec, Raleigh NC), c-met (3D4, dilution 1:100; Zymed, San Francisco, CA), and HGF (polyclonal, dilution 1:20; R&D Systems, Minneapolis, MN). Heat-induced epitope retrieval was achieved for all target antigens by heating sections in 0.1 mol/l of citrate buffer (ph 6.0) in a commercial microwave oven. Localization was via avidin biotin or streptavidin biotin immunoperoxidase method using 3,3'-diaminobenzadine-4HCl as the chromogen. Appropriate positive and negative control experiments were performed. The percentages of ER- and PR-positive stromal cells were semiquantitatively graded from 0 to 4 (0, 0% of stromal cells positive; 1, 1%-24% of stromal cells positive; 2, 25%- 49% of stromal cells positive; 3, 50%-74% of stromal cells positive; and 4, 75%-100% of stromal cells positive). The staining pattern and intensity of α-inhibin, c-met, and HGF were also semiquantitatively graded from 0 to 4 (0, no staining; 1, focal and weak; 2, diffuse and weak; 3, focal and strong; and 4, diffuse and strong). Statistical Analysis Data are expressed as medians, means ± SD, and percentages except as noted. One-way analysis of variance was used to evaluate for between-group differences in continuous data. Correction for multiple comparisons was made using the Bonferroni method. Between-group comparisons for semiquantitative data were made using the Kruskal-Wallis test and Wilcoxon rank sum test. All statistical analyses were performed using Stata 8.0 (Stata, College Station, TX), and the type I error level (α) was.05. Results The clinicopathologic features of all study cases are summarized in Table 1. All 8 patients who had HBCs were women between the ages of 31 and 64 years (mean, 45.7 years). Similarly, all 10 cases of PMCNs were also all found in women between the ages of 24 and 84 years (mean, 54.9 years). There was no statistically significant difference in age between patients with HBC and PMCN (P >.05). The 6 simple liver cysts were in 2 men and 4 women. The mean age of patients with simple liver cysts was older (62.1 years) than of patients with HBC (P <.05). However, there was no statistically significant difference in mean age between patients with PMCN and simple liver cyst. All 8 HBCs were found within the liver. Grossly, all were solitary and multilocular, each cavity containing fluid of variable consistency. No luminal communication with the surrounding bile ducts or pancreatic ductal system was identified in any case of HBC or PMCN, respectively, grossly or microscopically. Cysts of HBCs and PMCNs were lined by tall, columnar cells arranged in a single, flat row with occasional papillary or polypoid projections, pseudostratifications, and crypt-like invaginations. The columnar cells contained histologically benign-appearing, basally located nuclei and abundant intracellular mucin. Mild epithelial atypia with a slight increase in the size of the basally located nuclei was identified at least focally in all cases of HBC and PMCN. No intestinal metaplasia was identified. Mitotic activity was absent in all cases. The 212 Am J Clin Pathol 2008;129: DOI: /U2BBP4EMBAHCM6E6

3 Anatomic Pathology / ORIGINAL ARTICLE Table 1 Clinical, Histologic, and Immunohistochemical Results for Hepatobiliary Cystadenomas, Pancreatic Mucinous Cystic Neoplasms, and Simple Liver Cysts * Sex Stromal Age (y) (M/F) Density ER PR Inhibin HGF c-met Hepatobiliary mucinous 45.7 (11.09) 0:8 3.0 (0) 3.5 (1.07) 3.5 (0.76) 2.8 (0.71) 1.0 (0.76) 3.9 (0.35) cystadenoma (n = 8) Pancreatic mucinous 54.9 (18.56) 0: (0.92) 1.2 (1.40) 0.9 (1.29) 1.4 (1.43) 0.6 (0.52) 3.7 (0.67) cystic neoplasms (n = 10) Simple liver cysts (n = 6) 62.1 (5.42) 2:4 0.0 (0) 0.0 (0) 0.0 (0) 0.0 (0) 0.3 (0.82) 2.7 (1.03) ER, estrogen receptor; HGF, hepatocyte growth factor; PR, progesterone receptor. * Age and scores for stromal density and the markers are given as mean (SD). Grading was as follows: stromal density: 0, none; 1, scant; 2, intermediate; and 3, abundant; ER and PR: 1, 1%-24% cells positive; 2, 25%-49% cells positive; 3, 50%-74% cells positive; and 4, 75%-100% cells positive; inhibin, HGF, and c-met: 1, focal, weak; 2, diffuse, weak; 3, focal, strong; and 4, diffuse, strong. P >.05 between HBCs and PMCNs; P <.05 between HBCs and simple liver cysts. P <.05 for comparisons of all 3 groups. P >.05 for comparisons of all 3 groups. P <.05 when HBCs and simple liver cysts and when PMCNs and simple liver cysts were compared; P >.05 between HBCs and PMCNs. subepithelial stroma of HBCs and PMCNs consisted of a small inner layer of moderately to densely cellular mesenchyme overlying a dense layer of collagenous connective tissue. The former mesenchymal layer of both types of mucinous cyst resembled ovarian stroma, consisting of a compact arrangement of bland, spindle-shaped cells with round to oval nuclei. All 8 HBCs had abundant stroma Image 1A, whereas the stromal density of PMCNs was more variable Image 1B with rare to intermediate cellularity (mean stromal grade, 3.0 vs 1.8, respectively; P <.05). Subset analysis of age and likely hormonal status with HBC stromal density (data not shown) did not show a difference, as all of the HBCs had abundant stroma regardless of patient age. Although subset analysis of PMCN stromal density was too small for statistical review, there was a trend indicating that premenopausal women (<40 years, n = 2) had abundant stroma, whereas perimenopausal and menopausal women ( 40 years, n = 8) had rare to intermediate amounts of stroma (mean stromal grade, 3.0 vs 1.5, respectively; P <.05). Of note, although the average age in the HBC group tended to be younger, there was no significant difference in age between the HBC and PMCN groups. Simple liver cysts were lined by tall columnar cells arranged in a single, flat row but lacked the ovarian-like mesenchymal stroma. ER+ stromal cells were more abundant in HBCs than in PMCNs Image 1C and Image 1D (3.5 vs 1.2, respectively; P <.05). PR staining was also greater in HBCs than in PMCNs Image 1E and Image 1F (3.5 vs 0.9, respectively; P <.05). Focal strong reactivity for α-inhibin was consistently displayed in HBCs, whereas PMCNs typically had focal to diffuse, weak staining Image 1G and Image 1H (2.8 vs 1.4, respectively; P <.05). Stroma supporting simple liver cysts was negative for ER, PR, and α-inhibin. HGF was not expressed in the stromal cells of PMCNs or HBCs and was focal and weak (and not significantly different) in the epithelium of HBCs, PMCNs, and simple liver cysts Image 2A, Image 2B, and Image 2C. In contrast, c-met reactivity in the epithelium of HBCs and PMCNs was significantly stronger than that detected in simple liver cysts Image 2D, Image 2E, and Image 2F (3.9, 3.7, and 2.7, respectively; P <.05). There was no immunostaining for c-met in the adjacent hepatic and pancreatic parenchyma or pancreaticobiliary epithelium. Discussion Edmonson 17 defined HBC as a multilocular lesion lined by columnar epithelium with an accompanying densely cellular ovarian-like stroma. Devaney et al 2 regarded this definition as rather restrictive and, therefore, expanded the criteria to include cystadenoma with and without an ovarian-like stroma and having a multilocular or a unilocular gross appearance. A recent study examined biliary cystic tumors with bile duct communication. These cystic tumors had prominent papillary proliferation with fibrovascular cores while ovarian-like stroma was not observed, 1 reminiscent of the intraductal papillary neoplasm of the biliary tree with prominent cystic dilatation. These observations suggested that ovarian-type stroma and lack of communication with bile ducts are important diagnostic elements in HBCs. In the present study, all 8 HBCs were multilocular, had ovarian-like stroma, and did not show evidence of communication with the bile ducts. Distinction of simple liver cysts from HBCs did not seem to pose diagnostic problems in the present study because simple liver cysts are typically unilocular 2 and do not have ovarian-like stroma. PMCNs share fundamental histologic features with HBCs, including epithelial morphologic features and ovarian-like stroma. Although the clinicopathologic features of PMCNs and HBCs have been Am J Clin Pathol 2008;129: DOI: /U2BBP4EMBAHCM6E6 213

4 Lam et al / OVARIAN-TYPE STROMA IN LIVER AND PANCREATIC LESIONS B C D Image 1 A, Hepatobiliary cystadenomas (HBCs) had abundant stroma (H&E, 100). B, Stromal density in pancreatic mucinous cystic neoplasms (PMCNs) was more variable with rare to intermediate cellularity (H&E, 100). C and D, Stromal staining for estrogen receptor in HBCs (C, 200) was greater than in PMCNs (D, 200). extensively studied in isolation, careful immunohistochemical comparison of these lesions has not been previously reported, and the present study readdresses this oversight and focuses on immunohistochemical analysis of the ovarian-like stroma and considers the possible role of the HGF/c-met pathway in the pathogenesis of these cystic lesions. HGF (also known as scatter factor) and its receptor, the tyrosine kinase c-met, comprise a mitogenic, morphogenic, cytotoxic, and angiogenic pathway that is found in a variety of normal and neoplastic conditions in lungs, kidneys, thyroid, liver, pancreas, and ovary In the liver, HGF and c-met are elevated in hepatocellular carcinoma when compared with nontumor tissue.15,23 Similarly, c-met is detected in low levels in normal human exocrine pancreas but is up-regulated in Am J Clin Pathol 2008;129: DOI: /U2BBP4EMBAHCM6E6 pancreatic cancer cell lines.21 HGF is a paracrine mediator produced by mesenchymal cells that stimulates various epithelial cells It is of particular interest in the present context because it is a mitogenic24 and motogenic factor25 that also influences cell differentiation and the spatial cell arrangement of cell layers.26,27 Together, HGF and c-met are a primarily paracrine signaling mechanism for hepatocarcinogenesis15,28 and for the proliferation of intrahepatic biliary epithelial cells14 and various extrahepatic tissues, such as ovarian surface epithelial neoplasms.16 In the present study, HGF was not significantly elevated in HBCs or PMCNs, again when compared with simple liver cysts, but c-met was significantly elevated in HBCs and PMCNs when compared with simple liver cysts. These findings A

5 Anatomic Pathology / ORIGINAL ARTICLE F G H E and F, Stromal staining for progesterone receptor in HBCs (E, 200) was greater than in PMCNs (F, 200). G and H, There was more consistent, focal, strong reactivity to inhibin in HBCs (G, 200) than in PMCNs (H, 200). suggest that another c-met ligand may contribute to this mitogenic/motogenic pathway. We cannot exclude the possibility that the increased expression of c-met in the epithelial cells of HBCs and PMCNs is an epiphenomenon of neoplastic transformation in cells that gives rise to HBCs and PMCNs; however, induced overexpression of c-met by other existing stromal ligands remains a more likely explanation, one that may in part explain a common pathogenesis. It would be interesting to explore c-met expression in pancreatic intraepithelial neoplasms, intraductal papillary mucinous neoplasm of pancreas, intraductal papillary neoplasms of the bile ducts, and invasive tumors associated with each because these neoplasms do not have ovarian-like stroma. In previous analyses of HBCs and PMCNs2,5,6 and as confirmed in the present study, both lesions display a striking prevalence in middle-aged women. Further evidence of this apparent sex specificity is the immunohistochemical detection of ER, PR, and α-inhibin in subepithelial mesenchyme. Of interest, cystic nephroma and mixed epithelial and stromal tumor of kidney, which are under the recently proposed umbrella term, renal epithelial and stromal tumor, have similar clinical, histologic, and immunohistochemical manifestations.29 The presence of ovarian-type stroma and müllerian related immunohistochemical markers suggest that these tumors may originate from müllerian remnants misplaced during embryogenesis. Zamboni et al5 offer 2 developmental pathways that might account for these observations in HBCs and PMCNs: (1) Endodermally derived epithelium and primitive mesenchyme in the pancreas and liver may start proliferating Am J Clin Pathol 2008;129: DOI: /U2BBP4EMBAHCM6E E

6 Lam et al / OVARIAN-TYPE STROMA IN LIVER AND PANCREATIC LESIONS B C D E F A Image 2 A, B, and C, Epithelial expression of hepatocyte growth factor was focally weak and not significantly different between hepatobiliary cystadenomas (HBCs; A, 200) pancreatic mucinous cystic neoplasms (PMCNs; B, 200), and simple liver cysts (C, 200). D, E, and F, Epithelial c-met expression between HBCs (D, 200) and PMCNs (E, 200) was significantly stronger than in simple cysts (F, 200) Am J Clin Pathol 2008;129: DOI: /U2BBP4EMBAHCM6E6

7 Anatomic Pathology / ORIGINAL ARTICLE because of hypersensitization to female sex hormone stimulation. (2) Ectopic ovarian stroma in the pancreas and along the biliary tree may release hormones and growth factors that prompt endodermally derived epithelium in its vicinity to proliferate and form cystic neoplasia. Wheeler and Edmondson 3 noted that HBCs resemble embryonic gallbladder and large bile ducts. Each of the latter are derivatives of the foregut and consist of glandular endoderm surrounded by a primitive mesenchymal mesodermal layer. It is tempting to speculate that if the hepatic portion of the primitive foregut gives rise to HBCs, then other divisions of the foregut, like the pancreatic foregut, might produce similar tumors. 3 In this context, it is important to note that stroma of the fetal pancreas 30 and liver 31 expresses ER. The second theory offered by Zamboni and colleagues 5 that HBCs and PMCNs may originate from ectopic ovary incorporated in liver and pancreas is perhaps less plausible because the ovarian stroma seen in HBCs and PMCNs has never been reported to contain follicles. 32 On the other hand, because the pancreaticobiliary tree is within the embryologic pathway and germ cell descent, stroma without follicles cannot be dismissed. A final theory is that expression of ER, PR, and α-inhibin may be unrelated to tumorigenesis, instead merely a secondary phenomenon. The latter presumably accounts for ER expression by biliary epithelium in primary biliary cirrhosis. 33 Weihing et al 8 further suggest that once ER and PR are expressed in nascent HBCs and PMCNs, the hormonal milieu found in women may sustain their growth. Despite the morphologic similarities and possible shared common pathogenesis of HBCs and PMCNs, the present study reveals differences in stromal density between HBCs and PMCNs: all of the HBCs had dense and abundant stroma, whereas the stroma varied in density and abundance in PMCNs. Devaney et al 2 noted that the stroma in most HBCs (86%) consists of closely packed spindle cells. Our results are consistent with their findings. In the work by Zamboni and associates, 5 PMCNs that were rich in stroma were usually mucinous cystic adenomas and mucinous cystic borderline tumors. Tumors lacking an ovarian-like stroma were typically invasive mucinous cystadenocarcinomas. None of the HBCs or PMCNs in the present study showed borderline or invasive features, which, according to Zamboni et al, 5 likely explains the presence of dense and abundant stroma in our series of HBC. Interestingly, HBC typically exhibits more intense staining for ER, PR, and α-inhibin than PMCN, an observation that correlates with and, in fact, may be a result of higher stromal density in HBC 6-8,10 regardless of the age or hormonal status of the patient. The potential role of the differing hormonal milieu between the liver and the pancreas should also be considered; hepatocyte expression of estrogen and progesterone may in part drive receptor expression in tumor-associated stromal cells. The histologic and immunohistochemical findings in this study further support that HBCs and PMCNs are related neoplasms and may share a common pathogenic origin. These findings are sufficiently reliable to support the role of immunohistochemical analysis in the diagnosis of more challenging cystic neoplasms of the liver and pancreas. From the Department of Pathology, University of Washington School of Medicine, Seattle. Presented in part at the Annual Meeting of the United States and Canadian Academy of Pathology; March 1, 2005; San Antonio, TX. Address reprint requests to Dr Yeh: Dept of Pathology, University of Washington School of Medicine, 1959 NE Pacific St, NE140D, Box , Seattle, WA References 1. Zen Y, Fujii T, Itatsu K, et al. Biliary cystic tumors with bile duct communication: a cystic variant of intraductal papillary neoplasm of the bile duct. Mod Pathol. 2006;19: Devaney K, Goodman ZD, Ishak KG. Hepatobiliary cystadenoma and cystadenocarcinoma: a light microscopic and immunohistochemical study of 70 patients. Am J Surg Pathol. 1994;18: Wheeler DA, Edmondson HA. Cystadenoma with mesenchymal stroma (CMS) in the liver and bile ducts: a clinicopathologic study of 17 cases, 4 with malignant change. Cancer. 1985;56: Compagno J, Oertel J. Mucinous cystic neoplasms of the pancreas with overt and latent malignancy (cystadenocarcinoma and cystadenoma): a clinicopathologic study of 41 cases. Am J Clin Pathol. 1978;69: Zamboni G, Scarpa A, Bogina G, et al. Mucinous cystic tumors of the pancreas: clinicopathological features, prognosis, and relationship to other mucinous cystic tumors. Am J Surg Pathol. 1999;23: Thompson LD, Becker RC, Przygodzki RM, et al. Mucinous cystic neoplasm (mucinous cystadenocarcinoma of low-grade malignant potential) of the pancreas: a clinicopathological study of 130 cases. Am J Surg Pathol. 1999;23: McCluggage WG. Value of inhibin staining in gynecological pathology. Int J Gynecol Pathol. 2001;20: Weihing RR, Shintaku IP, Geller SA, et al. Hepatobiliary and pancreatic mucinous cystadenocarcinomas with mesenchymal stroma: analysis of estrogen receptors/progesterone receptors and expression of tumor-associated antigens. Mod Pathol. 1997;10: Ridder GJ, Maschek H, Flemming P, et al. Ovarian-like stroma in an invasive mucinous cystadenocarcinoma of the pancreas positive for inhibin: a hint concerning its possible histogenesis. Virchows Arch. 1998;432: Yeh MM, Tang LH, Wang S, et al. Inhibin expression in ovarian-type stroma in mucinous cystic neoplasms of the pancreas. Appl Immunohistochem Mol Morphol. 2004;12: Rishi M, Howard LN, Bratthauer GL, et al. Use of a monoclonal antibody against human inhibin as a marker for sex cord stromal tumors of the ovary. Am J Surg Pathol. 1997;21: Am J Clin Pathol 2008;129: DOI: /U2BBP4EMBAHCM6E6 217

8 Lam et al / OVARIAN-TYPE STROMA IN LIVER AND PANCREATIC LESIONS 12. Cooke I, O Brien M, Charnock FM, et al. Inhibin as a marker for ovarian cancer. Br J Cancer. 1995;71: Zheng W, Sung CJ, Hanna I, et al. Alpha and beta subunits of inhibin/activin as sex cord stromal differentiation markers. Int J Gynecol Pathol. 1997;16: Joplin R, Hishida T, Tsubouchi H, et al. Human intrahepatic biliary epithelial cells proliferate in vitro in response to human hepatocyte growth factor. J Clin Invest. 1992;90: Schoedel KE, Tyner VZ, Kim TH, et al. HGF, MET, and matrix-related proteases in hepatocellular carcinoma, fibrolamellar variant, cirrhotic and normal liver. Mod Pathol. 2003;16: Parrott JA, Skinner MK. Expression and action of hepatocyte growth factor in human and bovine normal ovarian surface epithelium and ovarian cancer. Biol Reprod. 2000;62: Edmonson HA. Tumors of the Liver and Intrahepatic Bile Ducts. Washington, DC: Armed Forces Institute of Pathology; Atlas of Tumor Pathology; First series, Fascicle Rubin JS, Chan AM, Bottaro DP, et al. A broad-spectrum human lung fibroblast-derived mitogen is a variant of hepatocyte growth factor. Proc Natl Acad Sci U S A. 1991;88: Kan M, Zhang GH, Zarnegar R, et al. Hepatocyte growth factor/hepatopoietin A stimulates the growth of rat kidney proximal tubule epithelial cells (RPTE), rat nonparenchymal liver cells, human melanoma cells, mouse keratinocytes and stimulates anchorage-independent growth of SV-40 transformed RPTE. Biochem Biophys Res Commun. 1991;174: Di Renzo MF, Olivero M, Ferro S, et al. Overexpression of the c-met/hgf receptor gene in human thyroid carcinomas. Oncogene. 1992;7: Di Renzo MF, Poulsom R, Olivero M, et al. Expression of the Met/hepatocyte growth factor receptor in human pancreatic cancer. Cancer Res. 1995;55: Di Renzo MF, Olivero M, Katsaros D, et al. Overexpression of the Met/HGF receptor in ovarian cancer. Int J Cancer. 1994;58: Ueki T, Fujimoto J, Suzuki T, et al. Expression of hepatocyte growth factor and its receptor c-met proto-oncogene in hepatocellular carcinoma. Hepatology. 1997;25: Miyazawa K, Tsubouchi H, Naka D, et al. Molecular cloning and sequence analysis of cdna for human hepatocyte growth factor. Biochem Biophys Res Commun. 1989;163: Stoker M, Gherardi E, Perryman M, et al. Scatter factor is a fibroblast-derived modulator of epithelial cell motility. Nature. 1987;327: Montesano R, Matsumoto K, Nakamura T, et al. Identification of a fibroblast-derived epithelial morphogen as hepatocyte growth factor. Cell. 1991;67: Montesano R, Schaller G, Orci L. Induction of epithelial tubular morphogenesis in vitro by fibroblast-derived soluble factors. Cell. 1991;66: Ljubimova JY, Petrovic LM, Wilson SE, et al. Expression of HGF, its receptor c-met, c-myc, and albumin in cirrhotic and neoplastic human liver tissue. J Histochem Cytochem. 1997;45: Turbiner J, Amin MB, Humphrey PA, et al. Cystic nephroma and mixed epithelial and stromal tumor of kidney: a detailed clinicopathologic analysis of 34 cases and proposal for renal epithelial and stromal tumor (REST) as a unifying term. Am J Surg Pathol. 2007;31: Greenway B, Iqbal MJ, Johnson PJ, et al. Oestrogen receptor proteins in malignant and fetal pancreas. Br Med J. 1981;283: Iqbal MJ, Wilkinson MI, Johnson PJ, et al. Sex steroid receptor proteins in foetal, adult and malignant human liver tissue. Br J Cancer. 1983;48: Clement PB, Young RH, Scully RE. Peritoneum. In: Sternberg SS, ed. Diagnostic Surgical Pathology. 2nd ed. New York, NY: Raven Press; 1994: Jorge AD, Stati AO, Roig LV, et al. Steroid receptors and heatshock proteins in patients with primary biliary cirrhosis. Hepatology. 1993;18: Am J Clin Pathol 2008;129: DOI: /U2BBP4EMBAHCM6E6