ANTICANCER RESEARCH 31: 3423-3428 (2011)

Clinical Value of Combined Detection of Serum Matrix Metalloproteinase-9, Heparanase, and Cathepsin for Determining Ovarian Cancer Invasion and Metastasis

WEI ZHANG1, HSIN-CHIH YANG1, QI WANG1, ZHI-JUN YANG1, HONG CHEN1, SU-MEI WANG1, ZHONG-MIAN PAN1, BU-JIAN TANG1, QINGDI QUENTIN LI2 and LI LI1

1Department of Gynecologic Oncology, Cancer Hospital, Guangxi Medical University, Nanning, Guangxi 530021, P.R. China; 2National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD 20892, U.S.A.

Abstract. Aim: This study evaluated the clinical value of groups. The serum CL, Hpa, and MMP-9 levels correlated the combined detection of serum cathepsin L (CL), with the degree of differentiation and the FIGO staging heparanase (Hpa), and matrix metalloproteinase-9 (MMP- (p>0.05). The serum CL (p=0.030) and MMP-9 (p=0.010) 9) for determining the degree of ovarian cancer invasion levels were significantly associated with peritoneal and metastasis before surgery. Patients and Methods: metastasis, and the serum Hpa level (p=0.042) was Enzyme-linked immunosorbent assays were used to measure associated with distant metastasis. A ROC curve analysis the serum content of CL, Hpa, and MMP-9 in 217 patients revealed sensitivity of 60.9%, 69.6%, and 72.2%, and with untreated ovarian cancer before surgery, 100 patients specificity of 57.4%, 67.2%, and 68.9% for the preoperative with benign ovarian tumors, and 101 healthy women as serum levels of CL, Hpa, and MMP-9, respectively, as tumor controls. In addition, the degrees of invasion and metastasis markers for the degree of extra-pelvic metastasis. were assessed by the ‘gold standard’ of clinicopathological Conclusion: Elevated serum CL, Hpa, and MMP-9 levels diagnosis. The associations of the preoperative serum CL, are correlated with malignant invasion and progression in Hpa, and MMP-9 levels with the clinicopathological factors ovarian cancer. The combined detection of serum CL, Hpa, and metastatic status were analyzed. Receiver operating and MMP-9 may be useful for determining the extent of characteristic (ROC) curve analysis was used to evaluate ovarian cancer metastasis before surgery. the usefulness of these markers for determining the degree of ovarian cancer invasion before surgery. Results: The Cytoreductive surgery combined with platinum-based serum CL, Hpa, and MMP-9 levels were significantly higher chemotherapy is the standard treatment for patients with (p=0.001) in patients with malignant ovarian cancer ovarian cancer; however, about 50-80% of patients relapse compared with patients with benign ovarian tumors and and/or gradually become drug-resistant within 18 to 28 healthy controls. The serum CL level was significantly months after the initial treatment (1-3). In most cases, death higher in patients with epithelial ovarian carcinoma is attributable to widespread tumor invasion and metastasis compared with non-epithelial ovarian carcinoma (p=0.048), to multiple organs. Thus, invasiveness and metastatic status whereas the serum levels of Hpa (p=0.109) and MMP-9 are life-threatening features of ovarian tumors. (p=0.544) did not differ significantly between these two Accurate preoperative assessments of the degree of malignancy and extent of metastasis are critical for optimal debulking, which is currently the best available approach in treating ovarian cancer. No tumor markers have been Correspondence to: Dr. Qingdi Quentin Li, National Institutes of established for evaluating ovarian cancer and other tumor Health, Building 10, Room 11N234, Bethesda, MD 20892-1888, metastasis, although some studies have suggested that the U.S.A. e-mail: [email protected] or Dr. Li Li, Guangxi overexpression of cathepsin L (CL), heparanase (Hpa), and Medical University Cancer Hospital, Nanning, GX 530021, P.R. matrix metalloproteinase-9 (MMP-9) may be associated with China. e-mail: [email protected] ovarian tumor metastasis (4-6). Key Words: Ovarian cancer, diagnosis, biomarkers, enzyme-linked As matrix hydrolases, CL, Hpa, and MMPs degrade immunosorbent assay, cathepsin, matrix metalloproteinase-9, various components of the extracellular matrix and play heparanase. important roles in many physiological processes and in

0250-7005/2011 $2.00+.40 3423 ANTICANCER RESEARCH 31: 3423-3423 (2011) tumors. In some types of cancer, the expression levels of CL, Table I. Regression equations, y=a(1-e–bx), for serum CL, Hpa, and Hpa, and certain MMPs in primary tumors can predict the MMP-9 levels as determined by ELISA. risk for metastasis (7). For example, elevated levels of CL, Marker a b r Hpa, and some MMPs were associated with accelerated tumor progression in colorectal cancer (8), bladder cancer (9), CL 1096.1137 0.0416 0.8578 hepatocellular carcinoma (10), breast cancer (11), and non- Hpa 1165.6651 0.0259 0.8398 small cell lung cancer (12). They have also been linked to the MMP-9 678.2558 0.9717 0.6769 aggressiveness of gynecological tumors, such as cervical CL, cathepsin L; Hpa, heparanase; MMP-9, matrix metalloproteinase- cancer (13) and ovarian carcinoma (14-17). CL, Hpa, and 9; ELISA, enzyme-linked immunosorbent assay. MMPs that are secreted from tumor cells can enter the blood stream, leading to increased serum levels of these proteins. Elevated serum or plasma levels of MMP-9 were observed in patients with colorectal or breast cancer (18, 19). The serum Methods. The serum concentrations of CL, Hpa, and MMP-9 were determined using enzyme-linked immunosorbent assay (ELISA) kits levels of CL, Hpa, and MMPs are shown to be related to their (Yingke Xinchuang BioTec, Xiamen, China). Goat polyclonal levels in tissues (20). Furthermore, serum levels of MMP-9 antibodies against CL, Hpa, and MMP-9, as well as other are significantly higher in patients with ovarian carcinomas antibodies, were purchased from Santa Cruz Biotechnology (Santa than in healthy women (21, 22). These findings suggest that Cruz, CA, USA). The optical densities were measured at 450 nm, the serum levels of CL, Hpa, and MMP may be useful as and the target protein concentrations in the samples were determined markers for evaluating the degree of malignancy and extent based on standard curves established using CurveExpert 1.3 of metastasis in ovarian cancer patients. An accurate software (Table I). preoperative assessment of these features would be beneficial Statistical data analyses. All data analyses, including t-tests and in determining the risks and timing of surgery. one-way ANOVA, were performed using SPSS 13.0 software The purpose of this study was to investigate the clinical (Statsoft, USA). Values of p<0.05 were taken to indicate statistical value of serum CL, Hpa, and MMP-9 levels as markers for significance. Receiver operating characteristic (ROC) curves were assessing the degree of malignancy and extent of metastasis analyzed to determine sensitivities and specificities. in ovarian cancer patients, by analyzing the correlation between the preoperative marker levels and the metastatic Results status of the patients. Serum CL, Hpa, and MMP-9 levels in the benign ovarian Patients and Methods tumor, malignant ovarian cancer, and healthy control groups. The serum CL, Hpa, and MMP-9 levels were much higher Patient samples. Patients in the Gynecologic Oncology Department, in the malignant ovarian cancer group compared with the Cancer Hospital, Guangxi Medical University, from September 2003 benign ovarian tumor and healthy control groups (p=0.001). to October 2009 were enrolled. The diagnoses were confirmed by Only serum CL level was higher in the benign ovarian tumor pathological findings after surgery. There were 217 cases of ovarian group compared with the healthy control group (p=0.001) malignancies, including 109 cases of serous cystic adenocarcinoma, 54 cases of mucous cystic adenocarcinoma, 14 cases of low or non- (Table II). differentiated carcinoma, 19 cases of sex cord stromal tumors, and 21 cases of malignant germ cell tumors. The surgical pathological Relationship between serum CL, Hpa, and MMP-9 levels and staging, determined in accordance with FIGO standards, showed 83 clinicopathological factors. In the malignant ovarian cancer stage I-II and 134 stage III-IV cases. The median age of the patients group, the serum CL level was significantly higher in cases in the malignant ovarian cancer group was 44.6 years (range, 16-67 of epithelial malignancy compared with non-epithelial years). There were 100 cases in the benign tumor group, including malignancy (21.598±8.249 vs. 19.565±7.698 ng/ml, 62 cases of serous cystic adenoma, 24 cases of mucous cystic adenoma, and 14 cases of benign teratoma; the median patient age respectively; p<0.05), but the CL level did not differ among was 35.6 years (14-64 years). As a control group, 101 healthy the different types of epithelial cancer (serous, mucous, and women, with a median age of 43.4 years (25-53 years), were others). The serum Hpa and MMP-9 levels were not recruited from a health screening. Before any treatment was begun, significantly different between epithelial and non-epithelial 2 ml of peripheral venous blood were collected. The serum was malignancies, or among different types of epithelial cancer separated by centrifugation at 3,000 rpm for 5 min and kept at (Table III). The serum CL, Hpa, and MMP-9 levels were –80˚C. This study was endorsed by the Ethics Committee of the each significantly associated with the FIGO stage and with Guangxi Medical University. All participants received an explanation of the aims of the study and signed a written informed the degree of differentiation (p<0.05 for each). consent; all understood that they could withdraw from the study at any time without influencing their oncological or general medical Relationship between serum CL, Hpa, and MMP-9 levels and treatment. ovarian cancer metastasis. None of the three serum markers

3424 Zhang et al: Assessment of Circulating Biomarkers for Metastatic Ovarian Cancer

Table II. Serum CL, Hpa, and MMP-9 levels in the healthy control, benign ovarian tumor, and ovarian malignancy groups.

Marker Group Cases Concentration (ng/ml)

(n) Median 25th percentile 75th percentile Range

CL Control 101 5.59±1.75 4.28 6.62 2.55-10.21 Benign 100 10.97±3.84 7.96 13.02 5.10-23.03 Malignancy 217 21.23±8.17 16.42 24.91 5.92-52.81 Hpa Control 101 2.77±0.80 2.17 3.12 1.54-5.82 Benign 101 4.86±1.37 3.88 5.82 2.11-9.15 Malignancy 217 7.68±2.34 6.28 9.25 2.39-13.31 MMP-9 Control 101 57.99±11.42 50.12 63.11 39.04-107.18 Benign 101 143.66±28.47 121.97 168.55 70.53-207.13 Malignancy 217 193.95±42.49 161.57 224.21 69.35-316.20

CL, cathepsin L; Hpa, heparanase; MMP-9, matrix metalloproteinase-9.

Table III. Serum CL, Hpa, and MMP-9 levels associated with Table IV. Serum CL, Hpa, and MMP-9 levels associated with ovarian clinicopathological factors in ovarian cancer. cancer metastasis.

Pathology Cases CL Hpa MMP-9 Metastasis Cases CL Hpa MMP-9 (n) (ng/ml) (ng/ml) (ng/ml) (n) (ng/ml) (ng/ml) (ng/ml)

Epithelial 177 21.598±8.249 8.054±2.05 195.743±41.667 Lymph node Serous 109 21.62±8.52 8.10±2.02 194.25±43.47 + 85 23.64±8.89 8.11±2.09 194.95±43.54 Mucous 54 20.28±7.44 7.97±2.14 195.84±37.83 − 92 21.42±8.82 8.00±2.04 196.47±40.08 Other* 14 26.49±7.64 8.06±2.17 207.04±42.62 Pelvic Non-epithelial 40 19.565±7.698 6.02±2.27 186.019±45.671 + 156 21.91±8.35 8.15±2.10 197.89±42.63 Degree of − 21 19.13±7.14 7.34±1.61 179.77±29.87 differentiation Peritoneal High-medium 29 18.54±7.30 7.20±2.51 173.43±39.37 + 115 22.96±8.41 8.20±2.19 200.44±43.82 Low 148 23.04±7.67 8.22±1.92 200.12±40.82 − 62 19.07±7.36 7.29±1.76 181.04±36.10 FIGO stage Distant I-II 62 19.66±7.83 7.21±2.05 182.63±42.30 + 32 22.03±8.05 8.63±1.69 201.74±41.07 III-IV 115 22.64±8.31 8.51±1.92 202.81±39.74 − 143 21.50±8.32 7.93±2.11 194.42±41.82

*Including endometrioid, clear-cell, undifferentiated, and other rare CL, cathepsin L; Hpa, heparanase; MMP-9, matrix metalloproteinase-9. pathological types. CL, cathepsin L; Hpa, heparanase; MMP-9, matrix metalloproteinase-9.

differed significantly between positive and negative status for Table V. Area under the ROC curve for serum CL, Hpa, and MMP-9 lymph node metastasis or for pelvic metastasis (p>0.05). The levels in ovarian cancer. serum CL and MMP-9 levels were higher in patients who Marker Area Standard error P-value 95% Confidence interval were positive for peritoneal metastasis, and the serum Hpa level was higher in patients with distant metastasis (p<0.05) CL 0.624 0.045 0.007 0.535-0.712 (Table IV). Hpa 0.685 0.044 0.000 0.599-0.772 MMP-9 0.746 0.038 0.000 0.671-0.820 Clinical value of the serum enzyme markers for assessing CA125 0.526 0.043 0.572 0.441-0.611 ovarian cancer metastasis. The ROC curves were established CA, cancer antigen; CL, cathepsin L; Hpa, heparanase; MMP-9, matrix based on the serum levels of CL, Hpa, MMP-9, and an metalloproteinase-9; ROC, receiver operating characteristic. established ovarian cancer biomarker, cancer antigen 125 (CA125), in 177 patients with epithelial ovarian cancer (Figure 1 and Table V). The area under the curve was CL, Hpa, and MMP-9 levels are potential markers for the compared between 115 cases with pelvic metastasis and 62 preoperative assessment of tumor metastasis in ovarian cases without, to estimate the sensitivities and specificities cancer; of these three, MMP-9 appears to be the best of the markers (Table VI). The results suggest that the serum indicator of metastasis (Table VI).

3425 ANTICANCER RESEARCH 31: 3423-3423 (2011)

Table VI. Predictive value of serum CL, Hpa, and MMP-9 levels for tumor metastasis in ovarian cancer.

Marker ROC Sensitivity Specificity False negative False positive Laminin receptor- Laminin receptor- area (%) (%) (%) (%) positive negative

CL 0.624 60.9 57.4 39.1 43.6 1.4 0.7 Hpa 0.685 69.6 67.2 31.4 32.8 2.1 0.5 MMP-9 0.746 72.2 68.9 27.8 31.1 2.3 0.4 CA125 0.526 – – – – – –

CA, cancer antigen; CL, cathepsin L; Hpa, heparanase; MMP-9, matrix metalloproteinase-9; ROC, receiver operating characteristic.

Discussion stage, differentiated and epithelial type ovarian carcinoma. These results are consistent with the proposed molecular Pelvic implantation and direct infiltration represent the major mechanism of cancer cell infiltration and metastasis, as well pathway of metastasis in ovarian cancer development, as our previous report on CL, Hpa, and MMP-9 expression compared with the less common lymph node metastasis and in ovarian cancer tissues and peripheral blood (28). even rarer blood pathway (23). Matrix-degrading enzymes, In addition, our results demonstrated further elevation of such as MMPs, CL, and Hpa, are essential in the process of serum CL and MMP-9 in ovarian cancer patients with tumor cell shedding, implantation, and infiltration, and the peritoneal metastasis, and increased serum Hpa in ovarian degree of tumor cell invasiveness is closely associated with cancer patients with distant metastasis, in general agreement the production or induction of these enzymes (24). Within 4- with earlier studies (29). The differences in the correlations 8 h after cancer cells contact the basement membrane, with metastasis sites may be attributable to differences in the enzymes produced by the tumor cells or by nearby interstitial detection methods or among individual samples, and to or infiltrated immune cells begin breaking down extracellular inaccurate estimation of multiple metastases. Sakata and matrix components, such as laminin, fibronectin, colleagues (30) discovered higher MMP-9 expression in proteoglycan, and collagenase IV. The cancer cells migrate ovarian cancer tissues with lymph node metastasis than without through local defects created in the basement membrane and lymph node metastasis, and proposed an association between infiltrate healthy tissues. high MMP-9 expression levels and lymph node metastasis. In the present study, serum levels of CL, Hpa, and MMP- Zhongxian et al. (31) reported a relationship between increased 9 were elevated in patients with malignant ovarian tumors cathepsin B expression and higher FIGO stage, metastasis, compared with patients having benign ovarian tumors and ascites, and a lower survival rate in ovarian cancer. Kodama healthy women (p<0.05); the CL level was also higher in and coworkers (32) have confirmed that ovarian tumor patients patients with benign tumors compared with healthy women with a large amount of ascites had higher Hpa expression (p<0.05). These findings are consistent with previous compared with patients having a small amount of ascites. reports. Manenti et al. (21) reported an increased MMP-9 All of the above studies link CL, Hpa, and MMP-9 with level in the peripheral blood of patients with ovarian cancer, metastatic ovarian cancer, and these associations may involve compared with healthy women and patients with benign one or more of the following possible mechanisms (33-35). ovarian tumors, suggesting the potential value of MMP-9 in Firstly, acting as proteases, these enzymes may break down diagnosing ovarian malignancy. The results presented by extracellular-matrix molecules such as collagen, laminin, and Karihtala and colleagues (25) also support MMPs as a fibronectin, thereby directly disintegrating the physical marker for monitoring chemotherapy in ovarian cancer. barrier to cancer cells and indirectly damaging the vessel Shinyo et al. (26) have shown an elevated serum Hpa level in basement membrane. Secondly, the disruption of the matrix patients with ovarian cancer, a drop in serum Hpa after tumor protein structure by these enzymes may rearrange resection, and an increase with relapse, suggesting a intercellular adhesion sites, thus facilitating the outward relationship between the Hpa level and tumor growth. growth of cancer cells. Thirdly, the proteolysis products of Siewinski and coworkers (27) reported that the serum CL matrix proteins may induce biological changes; for example, level is increased in patients with ovarian cancer. the digestion of laminin-5 by MMPs produces soluble It should be emphasized that our results suggest the chemotactic fragments that participate in the tumor immune involvement of all three of these enzymes during cancer response, and similar fragments produced by Hpa can inhibit infiltration and metastasis, as all were elevated concurrently activated T lymphocytes. Lastly, these enzymes may promote in ovarian malignancy. Moreover, the levels of CL, Hpa, and tumor proliferation and angiogenesis: CL may act as a MMP-9 were further elevated in patients with late FIGO mitogen to promote the proliferation of cancer cells; by

3426 Zhang et al: Assessment of Circulating Biomarkers for Metastatic Ovarian Cancer

metastasis in ovarian carcinoma patients are critical to achieve ideal reductive surgery and improve the 5-year survival rate. In the current study, we provide evidence that the serum levels of CL, Hpa, and MMP-9 were closely associated with ovarian cancer metastasis, exhibiting sensitivities of 60.9%, 69.6%, and 72.2%, respectively, and specificities of 57.4%, 67.2%, and 68.9%, respectively, as markers of ovarian cancer metastasis. MMP-9 had the most promising value. The combined detection of serum CL, MMP-9, and Hpa may indicate the extent of ovarian cancer metastasis before surgery. Further investigation of the combined detection of various markers in large-scale studies is warranted to establish their significance in ovarian cancer prognosis.

Acknowledgements

This study was supported by a grant from the Provincial Research Project Funding of Guangxi, China (No. GSR 9817101).

References

1 Deraco M, Baratti D and Laterza B: Advanced cytoreduction as surgical standard of care and hyperthermic intraperitoneal chemotherapy as promising treatment in epithelial ovarian cancer. Eur J Surg Oncol 37: 4-9, 2011. 2 Johnatty SE, Beesley J and Paul J: ABCB1 (MDR1) polymorphisms and progression-free survival among women with ovarian cancer following paclitaxel/carboplatin Figure 1. ROC curve analysis for CL, Hpa, MMP-9, and CA125 with chemotherapy. Clin Cancer Res 14: 5594-5601, 2008. regard to tumor metastasis in ovarian carcinoma. The ROC curves were 3 Chua TC, Liauw W and Robertson G: Second-line treatment of established based on the serum levels of CL, Hpa, MMP-9, and the first relapse recurrent ovarian cancer. Aust N Z J Obstet established ovarian cancer biomarker CA125 in 177 patients with Gynaecol 50: 465-471, 2010. epithelial ovarian cancer. The area under the curve was compared 4 Westermarck J and Kahari V-M: Regulation of matrix metallo- between 115 cases with pelvic metastasis and 62 cases without, to proteinase expression in tumour invasion. FASEB J 13: 781-792, estimate the sensitivities and specificities of the markers. The results of 1999. ROC curves suggest that serum CL, Hpa, and MMP-9 levels (area under 5 Nishikawa H, Ozaki Y and Nakanishi T: The role of cathepsin the curve=0.746) are potential markers for the preoperative assessment B and cystatin C in the mechanisms of invasion by ovarian of tumor metastasis in ovarian cancer; of these three, MMP-9 appears cancer. Gynecol Oncol 92: 881-886, 2004. to be the best indicator of metastasis. ca125, cancer antigen 125; cl, 6 Peretz T, Antebi SU and Beller U: Maintenance on extracellular cathepsin L; hpa, heparanase; mmp9, matrix metalloproteinase-9; ROC, matrix and expression of heparanase activity by human ovarian receiver operating characteristic. carcinoma cells from biopsy specimens. Int J Cancer 45: 1054- 1060, 1990. 7 Vihinen P and Kahari VM: Matrix metalloproteinases in cancer: prognostic markers and therapeutic targets. Int J Cancer 99: 157- remodeling the extracellular matrix, MMP-9 encourages the 166, 2002. 8 Kim TD, Song KS and Li G: Activity and expression of urokinase- growth of new blood vessels in tumors; and Hpa induces type plasminogen activator and matrix metalloproteinases in human budding and invasive growth of endothelial cells by releasing colorectal cancer. BMC Cancer 6: 211-221, 2006. heparin sulfate-bound growth factors such as basic fibroblast 9 Papathoma AS, Petraki C and Grigorakis A: Prognostic growth factor (bFGF) and vascular endothelial growth factor significance of matrix metalloproteinase-2 and -9 in bladder (VEGF) from the extracellular matrix. Furthermore, these cancer. Anticancer Res 20: 2009-2013, 2000. enzymes may act synergistically to promote angiogenesis. 10 Wang GB, Zhou XY and Wang XQ: Relationship between serum Although surgery is the major treatment option in ovarian heparanase and microscopic venous invasion in patients with hepatocellular carcinoma. Am J Clin Pathol 134: 242-248, 2010. cancer, its effect is often compromised by early and insidious 11 Theodoro TR, de Matos LL and Sant Anna AV: Heparanase extra-pelvic metastases such as subphrenic and mesenteric expression in circulating lymphocytes of breast cancer patients root lesions, especially in the superior abdomen region. The depends on the presence of the primary tumor and/or systemic early diagnosis and accurate preoperative assessment of metastasis. Neoplasia 9: 504-510, 2007.

3427 ANTICANCER RESEARCH 31: 3423-3423 (2011)

12 Naumnik W, Niklińska W and Ossolińska M: Serum cathepsin 24 Shield K, Ackland ML and Ahmed N: Multicellular spheroids in K and cystatin C concentration in patients with advanced non- ovarian cancer metastases: Biology and pathology. Gynecol small cell lung cancer during chemotherapy. Folia Histochem Oncol 113: 143-148, 2009. Cytobiol 47: 207-213, 2009. 25 Karihtala P, Mäenpää J and Turpeenniemi-Hujanen T: Front-line 13 Rauvala M, Aglund K and Puistola U: Matrix bevacizumab in serous epithelial ovarian cancer: biomarker metalloproteinases-2 and -9 in cervical cancer: different roles in analysis of the FINAVAST trial. Anticancer Res 30: 1001-1006, tumor progression. Int J Gynecol Cancer 16: 1297-1302, 2006. 2010. 14 Määttä M, Santala M and Soini Y: Matrix metalloproteinases-2 26 Shinyo Y, Kodama J and Hongo A: Heparanase expression is an and -9 and their tissue inhibitors in low malignant potential independent prognostic factor in patients with invasive cervical ovarian tumors. Tumour Biol 25: 188-192, 2004. cancer. Annals of Oncol 14: 1505-1510, 2003. 15 Brun JL, Cortez A, Commo F and Uzan S: Serous and mucinous 27 Siewinski M, Saleh Y and Gryboc M: Determination of cysteine ovarian tumors express different profiles of MMP-2, -7, -9, peptidase-like activity and their inhibitors in the serum of patients MT1-MMP, and TIMP-1 and -2. Int J Oncol 33: 1239-1246, with ovarian cancer treated by conventional chemotherapy and 2008. vitamin E. J Exp Ther Oncol 4: 189-193, 2004. 16 Ozalp S, Tanir HM and Yalcin OT: Prognostic value of matrix 28 Wang S, Li L and Zhang W: The clinical value of serum metalloproteinase-9 (gelatinase-B) expression in epithelial cathepsin L on diagnosis and monitoring of ovarian cancer. ovarian tumors. Eur J Gynaecol Oncol 24: 1417-1420, 2003. Zhongguo Zhongliu Linchuang 35: 167-168, 2008. 17 Nishikawa H, Ozaki Y and Nakanishi T: The role of cathepsin 29 Yurkovetsky Z, Skates S and Lomakin A: Development of a B and cystatin C in the mechanisms of invasion by ovarian multimarker assay for early detection of ovarian cancer. J Clin cancer. Gynecol Oncol 92: 881-886, 2004. Oncol 28: 2159-2166, 2010. 18 Somiari SB, Somiari RI and Heckman CM: Circulating MMP-2 30 Sakata K, Shigemasa K and Nagai N: Expression of matrix and MMP-9 in breast cancer potential role in classification of metalloproteinases (MMP-2, MMP-9, MT1-MMP) and their patients into low risk, high risk, benign disease and breast cancer inhibitors (TIMP-1, TIMP-2) in common epithelial tumors of the categories. Int J Cancer 119: 1403-1411, 2006. ovary. Int J Oncol 17: 673-681, 2000. 19 Hurst NG, Stocken DD and Wilson S: Elevated serum matrix 31 Zhongxian W, Shaoyang L and Daqiong J: The expression and metalloproteinase-9 (MMP-9) concentration predicts the significance of cathepsins in ovarian cancer tissues. Zhongliu 26: presence of colorectal neoplasia in symptomatic patients. Br J 559-562, 2006. Cancer 97: 971-977, 2007. 32 Kodama J, Shinyo Y and Hashen G: Heparanase messenger RNA 20 Wu ZS, Wu Q and Yang JH: Prognostic significance of MMP-9 expression in epithelial ovarian tumor. Int J Mol Med 12: 961- and TIMP-1 serum and tissue expression in breast cancer. Int J 964, 2003. Cancer 122: 2050-2056, 2008. 33 Wang M, Tang J and Liu S: Expression of cathepsinB and 21 Manenti L, Paganoni P and Floriani I: Expression levels of microvascular density increases with higher grade of vascular endothelial growth factor, matrix metalloproteinases-2 astrocytomas. J Neurooncol 71: 3-7, 2005. and -9 and tissue inhibitor of metalloproteinases-1 and -2 in the 34 He X, Brenchley PE and Janyson GC: Hypoxia increases plasma of patients with ovarian carcinoma. Eur J Cancer 39: heparanase-dependent tumor cell invasion, which can be 1948-1956, 2003. inhibited by antiheparanase antibodies. Cancer Res 64: 3928- 22 Downs LS, Lima PH and Bliss RL: Cathepsins B and D activity 3933, 2004. and activity ratios in normal ovaries, benign ovarian neoplasms, 35 Ricciardelli C and Rodgers RJ: Extracellular matrix of ovarian and epithelial ovarian cancer. J Soc Gynecol Investig 12: 539- tumors. Semin Reprod Med 24: 270-282, 2006. 544, 2005. 23 Fata JE, Ho AT and Leco KJ: Cellular turnover and extracellular matrix remodeling in female reproductive tissues: functions of Received May 28, 2011 metalloproteinases and their inhibitors. Cell Mol Life Sci 57: 77- Revised July 20, 2011 95, 2000. Accepted July 22, 2011

3428