Increased circulating hepatocyte growth factor (HGF): A marker of epithelial ovarian cancer and an indicator of poor prognosis

Gynecologic Oncology 121 (2011) 402–406

Contents lists available at ScienceDirect

Gynecologic Oncology j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y g y n o

Increased circulating hepatocyte growth factor (HGF): A marker of epithelial ovarian cancer and an indicator of poor prognosis Guro Aune a,⁎, Aina-Mari Lian b, Solveig Tingulstad c, Sverre H. Torp d, Siri Forsmo e, Janne Elin Reseland f, Astrid Kamilla Stunes a, Unni Syversen a,g a

Department of Cancer Research and Molecular Medicine, Faculty of Medicine, Norwegian University of Science and Technology (NTNU), Trondheim, Norway Oral Research Laboratory, Institute for Clinical Dentistry, University of Oslo (UiO), Oslo, Norway Department of Gynecological Oncology, Department of Laboratory Medicine, Children's and Women's Health, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway d Department of Pathology and Medical Genetics, Department of Laboratory Medicine, Children's and Women's Health, St. Olavs Hospital, Trondheim University Hospital, Trondheim, Norway e Department of Public Health and General Practice, Norwegian University of Science and Technology, Trondheim, Norway f Department of Biomaterials, Institute for Clinical Dentistry, University of Oslo (UiO), Oslo, Norway g Department of Endocrinology, St. Olavs Hospital, Trondheim University Hospital, Trondheim Norway b c

a r t i c l e

i n f o

Article history: Received 28 October 2010 Available online 1 February 2011 Keywords: Epithelial ovarian cancer Hepatocyte growth factor c-Met Hepatocyte growth factor activator

a b s t r a c t Objective. Hepatocyte growth factor (HGF) has been described to be increased in different cancers. In the present study we wanted to investigate whether HGF in serum can distinguish between benign and malignant ovarian tumors, and whether serum HGF levels can predict the outcome in patients with ovarian carcinomas. Methods. We included 123 consecutive patients appointed for laparotomy due to a pelvic mass. Preoperative levels of serum cancer antigen 125 (CA 125), HGF and HGF activator (HGFA) were quantified with immunological methods. We performed immunohistochemical analyses of HGFα, HGFβ and the receptor c-Met. Five-year survival of patients with advanced disease (stage III and stage IV) was analyzed with the Kaplan–Meier method. Results. Sixty patients had ovarian carcinomas, 23 borderline tumors, and 40 benign ovarian tumors. Patients with ovarian carcinomas had significantly higher preoperative HGF and CA 125 serum levels than patients with benign ovarian tumors, and borderline tumors. Patients with borderline tumors had significantly higher CA 125 values than benign cases. A combination of CA 125 and HGF increased the specificity in predicting carcinoma. We observed abundant HGFα, HGFβ and c-Met expressions in all ovarian tumors. Patients with advanced disease and preoperative serum HGF values ≥ 2 SD above reference value had a shorter disease-free survival than patients with advanced disease and serum HGF b 2 SD above reference value. Conclusions. HGF in serum is an indicator of ovarian carcinoma in women with a pelvic mass, and of a poor prognosis in advanced ovarian cancer. © 2011 Elsevier Inc. All rights reserved.

Introduction Ovarian cancer is the leading cause of death from gynecological malignancies in the Western world. Due to the late onset of symptoms, most patients have advanced disease at the time of diagnosis [1]. The prognosis is poor, with an overall survival-rate of about 40% in five years. So far, cancer antigen 125 (CA 125) is the only validated marker used in the treatment of ovarian cancer. It is, however, highly unspecific, and can be elevated in other conditions, both benign and malignant [2]. The identification of new tumor markers could contribute to the development of better diagnostic

⁎ Corresponding author. Fax: +47 72571463. E-mail address: [email protected] (G. Aune). 0090-8258/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.ygyno.2010.12.355

tools, improved diagnostic modalities, and tailored therapies for this malignancy. Hepatocyte growth factor (HGF) is a multifunctional growth factor, mainly produced by mesenchymal-derived cells. HGF is secreted in a pro-active single chain form, and is converted to the active heterodimeric form by proteolysis. One of the most significant proteases in the activation of HGF is the HGF activator (HGFA), a serine protease expressed by various tissues [3,4]. The effects of HGF are mediated by the tyrosine kinase receptor c-Met, predominately expressed by epithelial tissue [5–8]. Activation of the HGF/c-Met signaling pathway may lead to proliferation, enhanced cell motility, and angiogenesis. High levels of HGF and c-Met have been found in several human cancers [9–14]. C-Met is expressed in normal ovarian surface epithelium [7,15], and over-expressed in ovarian cancer [7,16,17]. High levels of HGF have been found in fluids of both benign and malignant ovarian cysts, as well as in ovarian cancer ascites [18,19].

G. Aune et al. / Gynecologic Oncology 121 (2011) 402–406

The aims of the present study were to investigate whether the serum levels of HGF can be used to distinguish benign from malignant ovarian tumors in women with a pelvic mass, and to evaluate whether preoperative serum HGF levels can predict the outcome in patients with ovarian carcinomas. Furthermore, we wanted to study the diagnostic and prognostic value of serum HGFA in epithelial ovarian cancer. Materials and methods Eligible for the study were women diagnosed with a pelvic mass appointed for laparotomy at the Gynecological Oncology unit at St. Olavs Hospital, the University Hospital of Trondheim, Norway. In the period from October 15th 2001 to April 30th 2005, 403 patients were referred due to a pelvic mass. We were able to collect blood samples from 184 of these patients, of whom 123 were included in this study. An informed consent was obtained from all participants. Patient files were reviewed to obtain data regarding age at diagnosis, histology, stage of disease, lymph node status, and follow-up. All patients were surgically staged according to the International Federation of Gynecology and Obstetrics (FIGO) staging system. Furthermore, all histologic slides were reviewed by one pathologist (S.H.T), and classified according to the World Health Organization guidelines as to histological type and grade of differentiation [20,21]. All patients were followed for five years after primary surgery or until death. The study was approved by the Regional Committee for Medical and Health Research Ethics. Serum analyses All serum samples were collected preoperatively. CA 125 was analyzed with immunological methods as a part of preoperative routine serum analyses, and the results were extracted from the patients' files. Serum was stored at − 80 °C until analyses. The serum levels of HGF were quantified using the Human Serum Adipokine, panel B kit (Linco Research, Inc., St. Charles, MI, USA) and the Luminex-100 system (Luminex Corporation, Austin, TX, USA). The fluorescence data were analyzed by the StarStation software (Version 2.0; Applied Cytometry Systems, Sheffield, United Kingdom). HGFA in serum was analyzed by an enzyme-linked immunosorbent assay (ELISA) (Human Activated HGFA Assay Kit — MCM, IBL, Gunma, Japan). The analyses were performed according to the manufacturers' instructions.

403

slides, de-paraffinized, and dehydrated. Antigen retrieval was performed by pressure cooking. The sections were incubated with a 1:50 dilution of HGFα and HGFβ-antibodies (IBL, Hamburg, Germany) for 40 min at room temperature, or a 1:200 dilution of c-Met antibody (Santa Cruz Biotechnology, Inc., CA, USA) overnight at 4 °C. Sections of breast carcinoma were used as positive controls for HGF, and sections of ventricular cancer were used as positive controls for c-Met. The separate controls were included in each separate staining performed. The Dako Envision kit (Dako Cytomation, Glostrup, Denmark) was used as secondary antibody, and an automatized immunohistostainer was applied for the analyses (Dako Techmate 500). Diaminobenzidine was used as chromogene and hematoxylin as counterstain. The primary antibody was omitted in the negative controls. The HGFα, HGFβ and c-Met immunoreactivities were scored as: −, negative; +, single positive tumor cells (b10%); ++, moderate fraction of positive tumor cells (10–50%); +++, high fraction of positive tumor cells (N75%); and ++++, very high fraction of positive tumor cells (N90%) [22]. Statistical analyses Statistical analyses were performed in the SPSS statistical software program 17.0. Graphs in Fig. 1 were made in Graph Pad Prism 5. For analyses of differences in serum levels between benign, borderline and ovarian carcinomas, we used the Kruskal–Wallis test. For subgroup analyses we used the Mann–Whitney U test. Analyses of correlations between the serum values of HGF, HGFA and CA 125 were performed with the Spearman correlation test. Correlation between age and serum HGF was also analyzed with the Spearman correlation test. Analyses of survival were carried out with the Kaplan and Meier method. Five-year overall survival was calculated from the date of primary surgery, to the date of death or date of censoring. Five-year disease free survival was calculated from the date of primary surgery, to the date of relapsed disease or date of censoring. P-values of b0.05 were considered statistically significant. Results The patients were diagnosed with malignant epithelial tumors, borderline tumors or benign ovarian tumors. Median age at diagnosis was 63.7 (37–83) years in the carcinoma group, 54.8 (21–83) years in the borderline group, and 58.9 (25–85) years in the benign group. Characteristics of the cases included are listed in Table 1.

Immunohistochemistry

Serum analyses

Formalin-fixed and paraffin-embedded sections were reviewed with a selection of representative sections for immunohistochemistry. Briefly, 4 μm-thick sections were mounted on Superfrost microscopic

The median serum HGF levels were 2673 (55–16,689) pg/ml in the carcinomas, 1668 (177–7955) pg/ml in the borderline tumors, and 1343 (74–12,643) pg/ml in the benign tumors (Fig. 1). Patients with

Fig. 1. Serum levels (median with range) of HGF, HGFA, and CA125 in patients with carcinomas, borderline tumors, and benign tumors of the ovary. *Comparison carcinomas/benign, significant differences, p b 0.05; ‡ comparison carcinomas/borderline, significant differences, p b 0.05; and † comparison borderline/benign, significant differences, p b 0.05.

404

G. Aune et al. / Gynecologic Oncology 121 (2011) 402–406

Table 1 Characteristics of the cases included. Characteristics

No. of cases

Carcinomas FIGO stage I II III IV Histologic type Serous Mucinous Endometroid Clear cell Mixed Undifferentiated Histologic grade (G)a G1 G2 G3 Borderline Serous Mucinous Clear cell Mixed Benign Epithelial/functional cyst Dermoid cyst Endometriosis Ovarian fibroma Uterine myoma

60

a

21 3 30 6 29 3 13 7 6 2 7 12 34 23 15 6 1 1 40 25 3 1 8 3

Clear cell tumors are not graded.

ovarian carcinomas had significantly higher preoperative HGF levels than patients with benign ovarian tumors (p = 0.001), and borderline tumors (p = 0.02). We found no significant serum HGF differences between borderline tumors and benign tumors (p = 0.38). The patients with advanced disease (stage III and stage IV) had significantly higher serum HGF values (3045 (511–16,689) pg/ml), compared to the patients with early stage disease (stage I and stage II; 1595 (55–5109) pg/ml) (p = 0.05). However, no differences between early stage disease and benign or borderline cases were detected (p = 0.2 and p = 0.6, respectively). We found no significant differences in serum HGF related to histological type (p = 0.37), lymph node metastasis (p = 0.85), or grade of differentiation (p = 0.6). To

evaluate a possible correlation between age and serum HGF values, patients were divided into four groups according to age; b30, 30–44, 45–54 and N54 years. There was no difference in preoperative serum HGF levels related to age (r = 0.119, p = 0.189). Serum HGFA was anlyzed in 50 carcinomas, 19 borderline tumors and 35 benign tumors. The median serum HGFA levels were 35.6 (6.7–60.8) pg/ml in the carcinomas, 43.4 (4.05–60.3) pg/ml in the borderline tumors and 39.2 (9.7–58.5) pg/ml in the benign tumors (Fig. 1). We did not find any differences in the HGFA concentration between the groups. However, serum HGFA levels showed a significant negative correlation to the levels of serum HGF (r = − 0.209, p = 0.032). The median CA 125 levels were 625 (29–15,000) KU/l in the carcinomas, 114 (7–2733) KU/l in the borderline tumors and 26 (5–1295) KU/l in the benign tumors (Fig. 1). The patients with carcinomas had higher preoperative CA 125 values compared to the patients with both benign tumors (p b 0.001) and borderline tumors (p b 0.001). Patients with borderline tumors had higher CA 125 values than benign cases (p = 0.002), and patients with advanced disease had higher CA 125 values than patients with early stage disease (p = 0.001). Furthermore, patients with early stage disease had higher serum CA 125 levels than both benign and borderline cases (p b 0.001 and p = 0.006, respectively). We found a significant positive correlation between the serum levels of CA 125 and HGF (r = 0.363, p b 0.001). In the analyses of the potential to predict malignancy, a cut-off value of 35 KU/l was determined for CA 125. The median serum HGF value in the benign group was chosen as reference HGF value, and a cut-off value was determined at 2 SD above this value; 5760 pg/ml (1343 + 4417). The HGF values under this cut-off are referred to as normal, whereas values above this cut-off are referred to as elevated. CA 125 had a sensitivity of 98.3% and a specificity of 59% in predicting ovarian carcinoma. When HGF was included in the analyses, there were no improvements in sensitivity. However, a combination of CA 125 and HGF increased the specificity to 94.8%. Immunohistochemistry Immunohistochemical analyses were carried out in 59 carcinomas, 11 borderline tumors and 11 cystadenomas (Table 1, Fig. 2). In all ovarian tumors, all over abundant HGFα, HGFβ and c-Met

Fig. 2. Results and images of the HGFα, HGFβ and c-Met immunohistochemical stainings in carcinomas, borderline and benign tumors of the ovary. i Results of the HGFα, HGFβ and c-Met immunohistochemical stainings in carcinomas, borderline tumors, and benign tumors of the ovary. ii Expression of HGFα, HGFβ, and c-Met in ovarian endometroid carcinoma (A),(B), (C) and serous cystadenoma (D),(E),(F), respectively. The images are from representative specimens.

G. Aune et al. / Gynecologic Oncology 121 (2011) 402–406

405

expressions were observed (Fig. 2i). The intensity varied from weak to strong, negative areas were, however, noticed as well (Fig. 2ii). Predominantly cytoplasmic immunoreactivity was found, in some cases membranous. Nuclear c-Met expression was also observed. Furthermore, the ovarian stromal cells were often weakly positive. In general, there were fewer positive cells, and the staining intensity was lower, in the mucinous tumors. The fallopian tubes also expressed HGFα and HGFβ. Survival analyses The most significant prognostic factor in our study was the stage of disease. Patients with early stage disease had longer overall and disease-free survival compared to patients with advanced stage disease (p b 0.001). Univariate analyses of prognostic factors for disease-free and overall survival in the cases of advanced disease are shown in Table 2. Elevated levels of serum HGF tended to imply a shorter overall survival. The median overall survival was 23 months in the cases with elevated serum HGF, as compared to 41 months in the cases with normal serum HGF (p = 0.052). We found a significant difference in disease-free 5-year survival in the cases of advanced disease, where the patients with normal HGF values had a longer disease-free survival, compared to the cases with elevated HGF (p = 0.037). Fig. 3 shows the Kaplan–Meier survival curve in cases of advanced disease, comparing cases with elevated HGF to cases with HGF in the normal range. We found no significant differences in survival related to the degree of immunohistochemical expressions of HGF α, HGF β and c-Met in the tumors. Discussion In the present study we found that patients with ovarian carcinomas had higher serum HGF levels than patients with borderline and benign ovarian tumors. This is in accordance with studies of HGF in several other cancers [9,11,12,14,23]. In order to evaluate the significance of serum HGF as a tumor marker, we compared the serum HGF levels with the levels of serum CA 125. The Table 2 Analyses of prognostic factors for 5-year disease-free and overall survival in the cases of stage III and stage IV ovarian carcinomas (Kaplan–Meier).

Prognostic variable Age b50 years ≥50 years Histology Serous Non-serous Gradea Low High Lymph node involvement + − Results of surgery Optimally debulked Suboptimally debulked Serum HGF b5760 pg/ml ≥5760 pg/ml Serum CA 125 35–100 KU/l N100 KU/l

n

Disease-free survival

Overall survival

%

p

%

p

2 34

0 5.6

N.S.

50 20.6

N.S.

24 12

0 16.7

0.02

16.7 33.3

N.S.

3 33

0 6.1

N.S.

33.3 21.2

N.S.

6 29

0 6.9

N.S.

33.3 20.7

N.S.

4 32

25.0 3.2

0.035

25.0 21.9

N.S.

21 15

9.5 0

0.037

28.6 13.3

N.S. (0.052)

1 35

0 5.7

N.S.

0 22.9

N.S.

N.S. = not significant (p ≥ 0.05). a Clear cell tumors are not graded.

Fig. 3. 5-year survival in advanced stage epithelial ovarian cancer in cases HGF ≤2 SD and N2 SD from HGF reference value.

serum CA 125 results in our study are in accordance with the previous studies [24–26]. In screening for early stage ovarian cancer, an efficient tumor marker requires a sensitivity N75%, and a specificity N99.5% [27]. We found a great increase in specificity, combining CA 125 and HGF, and this should be further elucidated. Elevated levels of serum HGF, as well as HGFA, have been described in cases of multiple myeloma compared to controls [28]. Nagakawa et al. found higher levels of serum HGF in prostate cancer compared to benign prostate hyperplasia, without corresponding elevations of HGFA [4]. This is in accordance with our results. In contrast to our results, however, Nagakawa et al. described a positive correlation between HGF and HGFA in the malignant cases. The significant negative correlation between HGF and HGFA we found could be due to negative feedback mechanisms, suppressing the production of HGFA in response to elevated HGF levels. To explore whether the serum HGF levels could derive from the ovaries, we examined the immunohistochemical expressions of HGFα, HGFβ and c-Met in the tumors. We found abundant expressions of HGFα, HGFβ and c-Met, indicating that the increased levels of circulating HGF may originate from ovarian tumor cells. One might question why the benign tumors expressed high immunoreactivity of both HGFα and HGFβ, without corresponding elevations in serum HGF. This could be caused by a better vascularization of the malignant tumors, leading to leakage of HGF into the circulation. We did not detect any differences in serum concentrations of HGF between the different histological types. However, in the immunohistochemical analyses, there were fewer positive cells, and the staining intensity was lower in the mucinous tumors. This could be caused by mucine, occupying the intracellular volume, thus leaving less vacant intracellular volume available for staining. Stage of disease is known as one of the most significant prognostic factors in ovarian cancer, and was also highly significant in the survival analyses in our study. We therefore analyzed survival in patients with advanced disease separately. We found that elevated preoperative serum HGF levels indicated a poor disease-free survival. This is in accordance with studies indicating prognostic impact of HGF in colorectal cancer, hepatocellular carcinoma and in myeloma [14,29,30]. What is the role of HGF in the development and progression of ovarian epithelial cancer? More than 90% of ovarian cancers originate in the epithelial surface of the ovary. Interactions between the epithelial surface and the underlying stroma play a role in the initiation and progression of ovarian tumors [31], and HGF could be an important signal substance in this interaction. HGF is mainly produced by mesenchymal cells and stimulates surrounding epithelial cells in a paracrine manner. Epithelial ovarian carcinomas often contain stromal

406

G. Aune et al. / Gynecologic Oncology 121 (2011) 402–406

components, and abnormal HGF production by these stromal cells might stimulate oncogenesis in a paracrine manner. In a study by Parrott et al., HGF was expressed at high levels by both normal ovarian surface epithelium, and by epithelial borderline and ovarian cancer cells, and to a lesser extent by ovarian stromal cells [31]. In our study, the ovarian stromal cells were often HGF positive, but to a much lesser extent than the surrounding cancer cells, thus supporting the results by Parrot et al. Ovarian surface epithelium cells are modified peritoneal mesothelial cells, that undergo a mesenchymal to epithelial cell transition during development, and this can explain why ovarian surface epithelium cells express both epithelial and mesenchymal characteristics. It is possible that the high amount of HGF produced by the ovarian surface epithelium could suppress the stromal HGF production. It seems likely that HGF production by ovarian cancer cells stimulates ovarian oncogenesis through autocrine mechanisms, thus explaining the poor diagnosis related to high HGF levels. Drugs that target receptor tyrosine kinases like the c-Met receptor have been introduced as a new class of antineoplastic therapeutics [32–34]. In ovarian cancer, blocking HGF effects has been shown to suppress cancer cell migration in vitro, to reduce ascites and tumor burden, and to increase survival in mice [35,36]. C-Met targeted therapies could limit the invasiveness and metastases of ovarian tumors, and should be further evaluated in the treatment of epithelial ovarian cancer. In conclusion, we found that HGF in serum is an indicator of ovarian carcinoma in women with a pelvic mass. Furthermore, elevated levels of serum HGF imply a poor prognosis in women with advanced epithelial ovarian cancer. Serum HGF and CA 125 in combination increase the diagnostic specificity compared to either marker alone. Conflict of interest statement No conflicts of interest are noted for any author.

References [1] Sherman ME, et al. Survival among women with borderline ovarian tumors and ovarian carcinoma: a population-based analysis. Cancer 2004;100(5):1045–52. [2] Stany MP, Maxwell GL, Rose GS. Clinical decision making using ovarian cancer risk assessment. AJR Am J Roentgenol 2010;194(2):337–42. [3] Kataoka H, Kawaguchi M. Hepatocyte growth factor activator (HGFA): pathophysiological functions in vivo. FEBS J 2010;277(10):2230–7. [4] Nagakawa O, et al. Serum hepatocyte growth factor activator (HGFA) in benign prostatic hyperplasia and prostate cancer. Eur Urol 2005;48(4):686–90. [5] Comoglio PM. Structure, biosynthesis and biochemical properties of the HGF receptor in normal and malignant cells. EXS 1993;65:131–65. [6] Bottaro DP, et al. Identification of the hepatocyte growth factor receptor as the c-Met proto-oncogene product. Science 1991;251(4995):802–4. [7] Di Renzo MF, et al. Overexpression of the Met/HGF receptor in ovarian cancer. Int J Cancer 1994;58(5):658–62. [8] Rosen EM, Nigam SK, Goldberg ID. Scatter factor and the c-Met receptor: a paradigm for mesenchymal/epithelial interaction. J Cell Biol 1994;127(6 Pt 2): 1783–7. [9] Tanaka K, et al. Circulating level of hepatocyte growth factor as a useful tumor marker in patients with early-stage gastric carcinoma. Scand J Gastroenterol 2004;39(8):754–60. [10] Maulik G, et al. Role of the hepatocyte growth factor receptor, c-Met, in oncogenesis and potential for therapeutic inhibition. Cytokine Growth Factor Rev 2002;13(1):41–59.

[11] Tanimoto S, et al. Prognostic significance of serum hepatocyte growth factor in clear cell renal cell carcinoma: comparison with serum vascular endothelial growth factor. J Med Investig 2008;55(1–2):106–11. [12] Taniguchi T, et al. Increase in the circulating level of hepatocyte growth factor in gastric cancer patients. Br J Cancer 1997;75(5):673–7. [13] Baykal C, et al. Overexpression of the c-Met/HGF receptor and its prognostic significance in uterine cervix carcinomas. Gynecol Oncol 2003;88(2):123–9. [14] Seidel C, et al. Elevated serum concentrations of hepatocyte growth factor in patients with multiple myeloma. The Nordic Myeloma Study Group. Blood 1998; 91(3):806–12. [15] Corps AN, Sowter HM, Smith SK. Hepatocyte growth factor stimulates motility, chemotaxis and mitogenesis in ovarian carcinoma cells expressing high levels of c-Met. Int J Cancer 1997;73(1):151–5. [16] Sawada K, et al. c-Met overexpression is a prognostic factor in ovarian cancer and an effective target for inhibition of peritoneal dissemination and invasion. Cancer Res 2007;67(4):1670–9. [17] Bu R, et al. HGF/c-Met pathway has a prominent role in mediating antiapoptotic signals through AKT in epithelial ovarian carcinoma. Lab Invest 2011;91(1):124–37. [18] Sowter HM, Corps AN, Smith SK. Hepatocyte growth factor (HGF) in ovarian epithelial tumour fluids stimulates the migration of ovarian carcinoma cells. Int J Cancer 1999;83(4):476–80. [19] Baykal C, et al. Comparison of HGF (hepatocyte growth factor) levels of epithelial ovarian cancer cyst fluids with benign ovarian cysts. Int J Gynecol Cancer 2003;13 (6):771–5. [20] Kommoss S, et al. Histological grading in a large series of advanced stage ovarian carcinomas by three widely used grading systems: consistent lack of prognostic significance. A translational research subprotocol of a prospective randomized phase III study (AGO-OVAR 3 protocol). Virchows Arch 2009;454(3):249–56. [21] Tavassoli, F.A., Devilee, P., ed. Pathology and genetics of tumours of the breast and female genital organs World Health Organization classification of tumours, ed. P. Kleihues, Sobin, L.H. Vol. 1. 2003, International Agency for Research on Cancer: Lyon. 432. [22] Reifenberger J, et al. Epidermal growth factor receptor expression in oligodendroglial tumors. Am J Pathol 1996;149(1):29–35. [23] Ren Y, et al. Hepatocyte growth factor promotes cancer cell migration and angiogenic factors expression: a prognostic marker of human esophageal squamous cell carcinomas. Clin Cancer Res 2005;11(17):6190–7. [24] But I, Gorisek B. Preoperative value of CA 125 as a reflection of tumor grade in epithelial ovarian cancer. Gynecol Oncol 1996;63(2):166–72. [25] Cooper BC, et al. Preoperative CA 125 levels: an independent prognostic factor for epithelial ovarian cancer. Obstet Gynecol 2002;100(1):59–64. [26] Vasilev SA, et al. Serum CA 125 levels in preoperative evaluation of pelvic masses. Obstet Gynecol 1988;71(5):751–6. [27] Bast Jr RC, et al. New tumor markers: CA125 and beyond. Int J Gynecol Cancer 2005;15(Suppl 3):274–81. [28] Wader KF, et al. Elevated serum concentrations of activated hepatocyte growth factor activator in patients with multiple myeloma. Eur J Haematol 2008;81(5): 380–3. [29] Toiyama Y, et al. Serum hepatocyte growth factor as a prognostic marker for stage II or III colorectal cancer patients. Int J Cancer 2009;125(7):1657–62. [30] Mizuguchi T, et al. Prognostic impact of surgical complications and preoperative serum hepatocyte growth factor in hepatocellular carcinoma patients after initial hepatectomy. J Gastrointest Surg 2009;13(2):325–33. [31] 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(3):491–500. [32] Christensen JG, Burrows J, Salgia R. c-Met as a target for human cancer and characterization of inhibitors for therapeutic intervention. Cancer Lett 2005;225 (1):1–26. [33] Wang MH, et al. Potential therapeutics specific to c-MET/RON receptor tyrosine kinases for molecular targeting in cancer therapy. Acta Pharmacol Sin 2010;31(9): 1181–8. [34] Liu X, Newton RC, Scherle PA. Developing c-MET pathway inhibitors for cancer therapy: progress and challenges. Trends Mol Med 2010;16(1):37–45. [35] Saga Y, et al. Expression of HGF/NK4 in ovarian cancer cells suppresses intraperitoneal dissemination and extends host survival. Gene Ther 2001;8(19): 1450–5. [36] Zillhardt M, Christensen JG, Lengyel E. An orally available small-molecule inhibitor of c-Met, PF-2341066, reduces tumor burden and metastasis in a preclinical model of ovarian cancer metastasis. Neoplasia 2010;12(1):1–10.