Prognostic Study on Tissue and Serum

Vol. 10, 4761–4768, July 15, 2004 Clinical Cancer Research 4761

Expression of Trypsinogen-1, Trypsinogen-2, and Tumor-Associated Trypsin Inhibitor in Ovarian Cancer: Prognostic Study on Tissue and Serum

Annukka Paju,1 Juhani Vartiainen,2 independent of stage and histological type of the tumor (P < Caj Haglund,3 Outi Itkonen,1 0.01). Kristina von Boguslawski,4 Arto Leminen,2 Conclusions: Tissue expression of TATI and an elevated 2 1 preoperative serum concentration of trypsin-2-API are Torsten Wahlstro¨m, and Ulf-Håkan Stenman strong independent prognostic factors in advanced epithelial Departments of 1Clinical Chemistry, 2Obstetrics and Gynecology, and 3 ovarian cancer. These results suggest that trypsin expression Surgery, Helsinki University Central Hospital, Helsinki, and plays a role in the progression of ovarian cancer. TATI and 4Department of Pathology, University of Helsinki, Helsinki, Finland trypsin-2-API are of potential use as an aid for stratification of randomized studies and for selecting treatment strategies. ABSTRACT Purpose: The purpose is to study the prognostic signif- INTRODUCTION icance of tissue expression of trypsinogen-1, trypsinogen-2, Epithelial ovarian carcinoma is a leading cause of death and tumor-associated trypsin inhibitor (TATI) and serum from gynecological malignancies in most developed countries. concentration of trypsinogen-2, trypsin-2-API (complex of The high mortality is largely because of advanced stage at trypsin-2 with ␣-1-proteinase inhibitor), and TATI in epi- presentation. Although stage, grade, and the presence of residual thelial ovarian cancer. disease correlate with survival, the individual patient outcome is Experimental Design: Expression of trypsinogen-1, not entirely predictable on the basis of these. Other clinical and trypsinogen-2, and TATI was determined by immunohisto- biological markers have therefore been studied in attempts to chemistry with monoclonal antibodies in tissue sections of identify additional prognostic factors (1). Tumor markers, espe- tumors from 119 patients with untreated primary epithelial cially CA 125 in serum, play an important role in the diagnosis ovarian cancer. Preoperative serum concentrations of tryp- and follow-up of ovarian cancer (2). Some other serum markers, sinogen-2, trypsin-2-API and TATI were analyzed using e.g., the ␤-subunit of human chorionic gonadotropin (3), inhibin specific immunoassays. (4), vascular endothelial growth factor (5), and human Results: Fifty-four percent of the tumors expressed kallikrein-6 (6) and kallikrein-10 (7) have been shown to have trypsinogen-1, 45% expressed trypsinogen-2, and 30% ex- prognostic value when measured before treatment. pressed TATI. In patients with stage III and IV disease, Invasion and metastasis of solid tumors requires proteolytic and elevated TATI enzymes that degrade the extracellular matrix and basement (0.002 ؍ TATI tissue expression (P -were associated with membranes (8). Therefore, factors affecting the proteolytic ac (0.048 ؍ concentration in serum (P adverse cancer-specific and progression-free survival in uni- tivity of cancer cells are potential predictors of survival, and variate analysis. In multivariate analysis, TATI tissue ex- tissue expression of matrix metalloproteinase (MMP)-2 (9), histological urokinase-type plasminogen activator (10), and plasminogen ,(0.0001 ؍ tumor grade (P ,(0.005 ؍ pression (P were independent activator inhibitor 1 (11) have been shown to be prognostic (0.0005 ؍ and stage (P ,(0.02 ؍ type (P prognostic factors for adverse cancer-specific survival and factors in ovarian cancer. Trypsinogen is a serine proteinase that may play a role in (0.0003 ؍ and grade (P (0.006 ؍ TATI tissue expression (P for progression-free survival. In multivariate analysis of all tumor invasion. It degrades a wide spectrum of extracellular patients and those with advanced disease, serum trypsin-2- matrix proteins (12) and activates proforms of other proteinases, API concentration was an adverse prognostic factor for including MMPs 1, 2, 8, 9, and 13 (13–16). Tumor-associated cancer-specific and progression-free survival, and it was trypsinogen-1 and trypsinogen-2 are strongly expressed in most ovarian cancers (17–19) and in several other cancers, e.g., pancreatic (20), gastric (21, 22) and colorectal cancer (23, 24), cholangiocarcinoma (25), and in esophageal squamous cell car- Received 2/3/03; revised 3/31/04; accepted 4/15/04. cinoma (26). In ovarian cancer, expression of trypsinogen is Grant support: Finnish Cancer Foundation, Finnish Academy of Sci- associated with tumor aggressiveness (17, 19), and in esopha- ences, Sigrid Juselius Foundation, Helsinki University Central Hospital, geal squamous cell carcinoma, it is strongly associated with and University of Helsinki. The costs of publication of this article were defrayed in part by the recurrence and poor prognosis (26). payment of page charges. This article must therefore be hereby marked Tumor-associated trypsin inhibitor (TATI) is a 6-kDa in- advertisement in accordance with 18 U.S.C. Section 1734 solely to hibitor of trypsin originally purified from urine of an ovarian indicate this fact. cancer patient and later found to be identical to pancreatic Requests for reprints: Annukka Paju, Department of Clinical Chemistry, Biomedicum Helsinki, Room A418a, Haartmaninkatu 8, PB 700, FIN- secretory trypsin inhibitor (27, 28). TATI is expressed together 00290 Helsinki, Finland. Phone: 358-9-4717-1725; Fax: 358-9-4717-1731; with trypsin in several tumors and cancer cell lines (29), and E-mail: [email protected]. elevated serum concentrations are common in several malignant

diseases (29). TATI is a useful serum marker for diagnosis of Table 1 Clinical characteristics of 119 ovarian tumors mucinous ovarian cancer and a prognostic factor for ovarian (30, Age (years) 31) and renal cell cancer (32). Mean 58 We have studied tissue expression of trypsinogen-1, Range 29–85 trypsinogen-2, and TATI by immunohistochemistry and ana- Stage ␣ I 49 (41%) lyzed the serum concentrations of trypsinogen-2, trypsin-2- -1- II 14 (12%) proteinase inhibitor (trypsin-2-API) and TATI in patients with III 47 (39%) epithelial ovarian cancer and compared the prognostic value of IV 9 (7%) these with other clinicopathological factors. Grade 1 57 (48%) 2 20 (17%) 3 39 (31%) PATIENTS AND METHODS Unknown 5 (4%) Patients. The study population consisted of 119 patients Histologic type treated for primary epithelial ovarian cancer at the Department Serous 65 (55%) of Obstetrics and Gynecology of Helsinki University Central Mucinous 21 (18%) Endometrioid 17 (14%) Hospital between 1990 and 1995. The mean age was 56 years Anaplastic 5 (4%) (range, 29–85 years). All tissue samples were obtained at pri- Mesonefroides 10 (5%) mary surgery, and serum samples were drawn Ͻ4 weeks before Mixed 1 (1%) surgery. Staging was done according to the 1988 International Tumor size (cm) Ͻ2 1 (1%) Federation of Gynecology and Obstetrics rules (33), with pelvic 2–10 37 (31%) and para-aortic lymphadenectomy whenever feasible. All pa- Ͼ10 81 (68%) tients, except those with stage IaG1 disease (n ϭ 19), received Residual tumor size (cm) cisplatinum-based combination chemotherapy as first-line treat- 0 76 (64%) Ͻ ment. Relapses were treated in casu. At the latest follow-up 0.5 5 (4%) 0.5–1 3 (3%) (January 12, 1999), 70 of 119 patients were alive with a median 1–2 8 (7%) follow-up time 60 months (range, 6–98 months). Among 49 Ͼ2 10 (8%) patients that had died, 46 died of ovarian cancer, 1 of another Unknown 1 (1%) cancer, and 2 of other diseases. Patient and tumor characteristics Palliation 4 (3%) Peritoneal carcinosis 12 (10%) are shown in Table 1. Ovarian cancer tissue obtained by surgery was frozen by immersion in liquid nitrogen and stored at Ϫ80°C until used for reverse transcriptase-PCR. The study has been performed in accordance with the principles of Declaration of incubations were carried out in moist chambers at room Helsinki. temperature. Between staining steps, the slides were rinsed in Antibodies and Immunoassays. A monoclonal antibody PBS. Peroxidase staining was visualized using 3-amino-9- predominantly reacting with trypsinogen-1 (18), MAB 1482, ethyl-carbazole [Sigma, A 5754; 0.2 mg/ml in 0.05 M acetate was from Chemicon International (Temecula, CA). Monoclonal buffer (pH 5.0)]. The sections were counterstained with Mayer’s antibodies specific for trypsinogen-2 (8F7) (34) and TATI (6E8) hemalum (Merck) and mounted with Aquamount (BDH). As (35) were prepared as described previously. TATI in serum was positive controls, paraffin sections of human pancreatic tissue measured by radioimmunoassay using reagents from Orion Di- were used. As negative controls, tissue sections were stained by agnostica (Oulunsalo, Finland) as described previously (27). A replacing the monoclonal primary antibodies with nonimmune cutoff value of 22 ␮g/liter was used (36). Trypsinogen-2 (34) mouse IgG. and trypsin-2-API (37) in serum were measured by time- All sections were reviewed and scored for statistical anal- resolved immunofluorometric assays as earlier described using ysis by an experienced gynecologic pathologist (T. Wahlstro¨m), cutoff values of 80 and 14 ␮g/liter, respectively. who was blinded to clinical and outcome data. The antigen Immunohistochemistry. Formalin-fixed, paraffin-embed- staining was categorized in four classes: negative (0), in which ded tissue sections (4-␮m) were deparaffinized in xylene and cases no staining of the neoplastic cells was seen; faintly posi- rehydrated in graded concentrations of ethanol to water, pre- tive (1), where a few neoplastic cells were faintly stained; treated with 0.4% pepsin (pH 1.8) for 30 min at 37°C, and unequivocally positive (2), where most but not all of the neo- treated with 0.3% hydrogen peroxide in methanol for 30 min to plastic cells were faintly to strongly stained; and strongly pos- quench endogenous peroxidase activity. Immunostaining was itive staining (3), designating that all neoplastic cells showed performed using the Elite ABC kit (Vectastain; Vector Labora- strong staining. In all cases, the nonneoplastic tissue in the tories, Burlingame, CA). Blocking serum was applied for 15 sections was completely negative. These findings were repro- min. The antibody against trypsinogen-1 was used at a dilution ducible in repeated experiments. of 1:1000 (ascites) and those against trypsinogen-2 and TATI at Reverse Transcriptase-PCR and Sequencing. Total concentrations 1 and 0.2 ␮g/ml, respectively. Antibodies were RNA was extracted from frozen ovarian cancer tissue using incubated on the tissue sections overnight at room temperature. RNeasy kit (Qiagen, Valencia, CA). The oligonucleotide prim- Both the biotinylated second layer antibody and the peroxidase- ers were constructed on the basis of published sequences for labeled avidin-biotin complex were incubated on the sections trypsinogen-1 and trypsinogen-2 (38) and TATI (39): 5Ј-AT- for 30 min. All dilutions were made in PBS (pH 7.0), and all GTTCTGTGTGGGCTTC-3Ј (trypsinogen sense) and 5Ј-TTG-

TAGACCTTGGTGTAGACTC-3Ј (trypsinogen antisense); and Finnzymes, Espoo, Finland], 0.25 mM of each deoxynucleoside 5Ј-TCAGCCATGAAGGTAACAG-3Ј (TATI sense) and 5Ј- triphosphate, 20 pmol of both sense and antisense primers, and CAAGGCCCAGATTTTTGA-3Ј (TATI antisense). The tryp- 1.6 units of Dynazyme DNA polymerase (Finnzymes). The sinogen primers produced a fragment of 156 bp and those for amplification conditions were as follows: 35 cycles at 95°C for TATI a fragment of 243 bp. Total RNA (1 ␮g) was transcribed 1 min and 55°C for 1 min (trypsinogen) and 40 cycles at 95°C into cDNA using SuperScript II-RT (Invitrogen-Life Technol- for 1 min and 53°C for 1 min (TATI). Water was used as a ogies, Inc., Paisley, United Kingdom) according to the manu- negative control and cDNA from COLO 205 colon adenocarci- facturer’s instructions. Contamination of RNA samples with noma cells as a positive control in all experiments. The agarose DNA was excluded with control reactions without reverse tran- gel electrophoresis and sequencing of the PCR products were scriptase. The reverse transcriptase product (1 ␮l) was amplified carried out as described previously (40). in a 40-␮l reaction volume in 1ϫ PCR buffer [10 mM Tris-HCl Statistical Analysis. Differences between groups were

(pH 8.8), 1.5 mM MgCl2,50mM KCl, and 0.1% Triton X-100; analyzed by the Mann-Whitney U test. Survival curves were

Fig. 1 A and B, immunohistochemical expression of tumor-associated trypsin inhibitor in stage III grade 3 serous cystadenocarcinoma. C and D, immunohistochemical expression of trypsinogen-1 and trypsinogen-2 (E) in stage I grade 1 serous cystadenocarcinoma. F, negative control. Scale bars, 100 ␮m(A and C) and 50 ␮m(B, D, E, and F).

constructed by the Kaplan-Meier method. Multivariate survival only stage was of prognostic significance (P Ͻ 0.0001). In analyses were performed by Cox regression. Statistical end univariate analysis of patients with stage III and IV tumors, high points were cancer-specific and progression-free survival meas- grade, stage, elevated preoperative TATI concentration in se- ured from the date of operation to the date of death, relapse, or rum, and positive tissue staining for TATI were associated with latest follow-up. Surviving patients were censored at the time of their last clinical control, and a case was censored if death resulted from unrelated disease. The relationship between tissue expression of TATI and concentration of trypsin-2-API in serum was analyzed by one-way ANOVA. All tests were two-sided, and the significance level was set to P of 0.05.

RESULTS TATI in tissue was detected as diffuse cytoplasmic and partially membrane-associated staining in 30% (36 of 119) of all tumors (Fig. 1, A and B). Expression was more frequent in mucinous (95%, 20 of 21) than in endometrioid (35%, 6 of 17), mesonefroid (20%, 2 of 10), and serous (15%, 7 of 48) cystadenocarcinomas. There was an inverse correlation between TATI expression and stage (P ϭ 0.0048) and grade (P ϭ 0.077). The expression of TATI in tissue correlated positively with its serum concentrations (P ϭ 0.035) and with tumor size (P ϭ 0.028). Trypsinogen-1 was expressed in 54% (64 of 119) (Fig. 1, C and D) and trypsinogen-2 in 45% (54 of 119) (Fig. 1E)of all tumors. The trypsinogen staining was apical and perinuclear granular corresponding to Golgi and secretory vesicle staining. Trypsinogen-1 and trypsinogen-2 expression was most frequent in mucinous (57 and 62%, respectively), serous (54 and 43%, respectively), and endometrioid (53 and 41%, respectively) cystadenocarcinomas. There were no significant correlations between trypsinogen-1 or trypsinogen-2 expression and stage, grade, or size of the tumor. Expression of trypsinogen-1 and trypsinogen-2 and TATI in ovarian cancer tissue was confirmed by reverse transcriptase-PCR and sequencing of the PCR products (data not shown). Serum TATI was elevated in 32 of 106 patients (30%). The median concentrations were significantly greater in patients with mucinous (25 ␮g/liter) than in those with serous tumors (15 ␮g/liter; P ϭ 0.004) and in patients with grade 3 (19 ␮g/liter) than grade 1 (13 ␮g/liter) tumors (P ϭ 0.03). Serum trypsinogen-2 was determined in 98 patients, and elevated levels oc- curred in 13 patients (13%). The median concentrations were significantly higher in patients with mucinous (49 ␮g/liter) than in those with serous (36 ␮g/liter) tumors (P ϭ 0.02) and in patients with stage IV (63 ␮g/liter) than stage I (34 ␮g/liter) disease (P ϭ 0.03). Preoperative trypsin-2-API concentration was determined in 53 patients, and elevated values were observed in 32 patients (60%). The concentrations were not related to histological type or stage, but they were significantly greater in patients with grade 3 (19 ␮g/liter) than in those with grade 1 (13 ␮g/liter) tumors (P ϭ 0.02). In univariate analysis of all patients, presence of ascites, Ͼ high grade, high stage, residual tumor size 2 cm, and lack of Fig. 2 A, Kaplan-Meier analysis of cancer-specific survival according strong positive staining for trypsinogen-1 were all associated to immunohistochemical expression of tumor-associated trypsin inhib- with short cancer-specific survival. An elevated preoperative itor (TATI) and Kaplan-Meier analysis (B) of cancer-specific survival serum trypsin-2-API concentration (Fig. 2) correlated with according to preoperative serum TATI concentration among patients with stage III and IV ovarian cancer. C, Kaplan-Meier analysis of progression-free survival (Table 2). In serous tumors, lack of progression-free survival according to preoperative serum trypsin-2-API strong positive staining for trypsinogen-2 was also associated concentration among patients with stage I, II, III, or IV epithelial with poor prognosis (P ϭ 0.03), whereas in mucinous tumors, ovarian cancer.

Table 2 Univariate analysis of 119 patients with ovarian cancer: relative risk of death according to age; ascites formation; grade; histologic type; residual tumor size; trypsinogen-1 tissue expression; and preoperative serum trypsin-2-API concentration Cancer-specific survival Progression-free survival n RRa 95% CI P RR 95% CI P Age (years) Յ60 66 1 1 Ͼ60 53 1.80 0.99–3.24 0.05 1.83 1.03–3.27 0.04 Ascites No 48 1 1 Yes 71 2.98 1.47–6.03 0.002 2.84 1.45–5.59 0.003 Grade 1–2971 1 3 39 7.35 3.85–14.0 Ͻ0.0001 7.03 3.77–13.1 Ͻ0.0001 Stage I–II 63 1 1 III–IV 56 4.48 2.30–8.71 Ͻ0.0001 4.51 2.37–8.58 Ͻ0.0001 Histological type Serous 65 1 0.96 1 0.82 Mucinous 21 1.02 0.59–1.76 0.95 0.98 0.57–1.70 0.95 Endometrioid 17 0.93 0.51–1.68 0.80 0.89 0.49–1.61 0.69 Residual tumor size Յ2cm 92 1 1 Ͼ2 cm 10 5.28 2.21–12.60 0.0002 5.56 2.37–13.04 0.0001 Trypsinogen-1 expression Negative or faint 80 1 1 Unequivocal or strong 39 0.48 0.23–0.99 Ͻ0.05 0.53 0.26–1.07 0.08 Serum trypsin-2-API Յ14 ␮g/liter 21 1 1 Ͼ14 ␮g/liter 32 2.88 0.95–8.71 0.06 3.58 1.20–10.66 0.02 a RR, relative risk; CI, confidence interval.

short cancer-specific and progression-free survival (Tables 3 trypsin-2-API in serum was related to grade, we performed a and 4, Fig. 2). Histological type was also of prognostic signif- multivariate analysis where only stage, histological type of the icance, patients with serous tumors having the best and those tumor, and trypsin-2-API were included. In this analysis, all with mucinous tumors the worst prognosis. factors were independent prognostic indicators for cancer- In multivariate analysis of all patients, stage and grade specific and progression-free survival (P values for trypsin-2- were independent prognostic factors for cancer-specific survival API are 0.004 and 0.001, respectively). In multivariate analysis (P ϭ 0.007 and 0.0001, respectively), but only grade was of stage III and IV tumors, TATI tissue expression (P ϭ 0.006), significant for progression-free survival (P ϭ 0.0001). Because grade (P Ͻ 0.0001), stage (P ϭ 0.01), and histological type

Table 3 Uni- and multivariate analysis of cancer-specific survival of patients with stage III or IV ovarian cancer: relative risk of death according to grade; histologic type; stage; TATI tissue expression; and preoperative serum TATI concentration Univariate analysis Multivariate analysis n RRa 95% CI P RR 95% CI P Grade 1–2281 1 3 27 5.64 2.52–12.66 Ͻ0.0001 12.4 4.02–38.5 Ͻ0.0001 Histologic type Serous 44 1 0.0001 1 0.01 Mucinous 4 3.54 1.51–8.28 0.004 1.20 0.31–4.70 0.79 Endometrioid 5 0.77 0.36–1.67 0.51 2.26 0.81–6.28 0.12 Stage III 47 1 1 IV 9 0.42 0.28–0.64 Ͻ0.0001 3.70 1.30–10.6 0.01 TATI expression Negative 48 1 1 Positive 8 3.98 1.68–9.48 0.002 7.67 1.79–32.8 0.006 Serum TATI Յ22 ␮g/liter 14 1 1 Ͼ22 ␮g/liter 35 2.18 1.01–4.70 0.048 0.96 0.40–2.30 0.92 a RR, relative risk; CI, confidence interval; TATI, tumor-associated trypsin inhibitor.

Table 4 Uni- and multivariate analysis of progression free survival of patients with stage III or IV ovarian cancer: relative risk of death according to grade; histologic type; stage; TATI tissue expression; and preoperative serum TATI concentration Univariate analysis Multivariate analysis n RRa 95% CI P RR 95% CI P Grade 1–2281 1 3 27 4.97 2.31–10.7 Ͻ0.0001 8.36 2.91–24.0 0.0001 Histologic type Serous 44 1 0.01 1 0.35 Mucinous 4 2.5 1.15–5.49 0.02 0.82 0.23–2.97 0.76 Endometrioid 5 0.84 0.39–1.78 0.64 1.71 0.66–4.40 0.27 Stage III 47 1 1 IV 9 1.87 1.24–2.85 0.003 2.04 0.76–5.50 0.16 TATI expression Negative 48 1 1 Positive 8 3.34 1.43–7.77 0.005 6.13 1.45–26.0 0.01 Serum TATI Յ22 ␮g/liter 14 1 1 Ͼ22 ␮g/liter 35 2.54 1.21–5.32 0.01 1.16 0.48–2.76 0.74 a RR, relative risk; CI, confidence interval; TATI, tumor-associated trypsin inhibitor.

(P ϭ 0.01) were independent prognostic factors for cancer- impaired survival. Possibly the lack of trypsinogen immunore- specific survival and TATI tissue expression (P ϭ 0.01) and activity in tumor tissue reflects rapid secretion and activation of grade (P ϭ 0.0001) for progression-free survival. Tissue expres- this protease. Activated trypsin is rapidly inactivated by ␣-2- sion of TATI was associated with a 6.1–7.7-fold increased risk macroglobulin. Serine proteases in complex with ␣-2-macro- of death (Tables 3 and 4). An elevated preoperative trypsin-2- globulin are not detected by antibodies and they cannot be API concentration in serum also predicted adverse cancer- measured in serum (44). Trypsin reaching circulation can also specific and progression-free survival independently of stage be inactivated by API, and we were able to measure this com- and histological type of the tumor (P ϭ 0.012 and 0.013, plex by a specific immunoassay (37). Sixty percent of the respectively), and this was associated with an 8.1-fold increase patients had elevated serum concentrations of trypsin-2-API in risk of death. Comparison of the tissue expression of TATI complex, and an elevated level was a strong prognostic factor with serum concentrations of trypsin-2-API showed that these were not correlated (ANOVA, P ϭ 0.9; Fig. 3).

DISCUSSION We have earlier shown that elevated preoperative serum TATI concentrations predict adverse prognosis in patients with stage III and IV epithelial ovarian cancer (3, 31). The results of the present study confirm these findings, and they additionally show that immunohistochemical expression of TATI also is a strong prognostic factor independent of tumor stage, grade, and histological type. Interestingly, the tissue expression of TATI was a much stronger prognostic factor than serum TATI, and its prognostic significance in advanced disease was actually com- parable with that of grade and stage. Because trypsin is thought to mediate tumor invasion (12, 16) and TATI is a trypsin inhibitor, the correlation between high TATI concentrations in serum and tissue and poor prognosis may seem surprising. However, a poor prognosis has also found to be associated with high levels of another proteinase inhibitor, plasminogen activator inhibitor 1, in breast (41), lung (42), and colorectal cancer tissue (43). The prognostic value of TATI has been explained by the association between TATI and trypsinogen expression. Thus, it Fig. 3 Comparison between immunohistochemical tissue expression of was surprising to find that tissue expression of trypsinogen-1 tumor-associated trypsin inhibitor (TATI; staining intensity 0–3) and serum concentrations of trypsin-2-API in samples from patients with and trypsinogen-2 was not associated with poor prognosis in any ovarian cancer. The dashed line shows the upper reference limit of subset of patients. In contrast, univariate analysis showed that serum trypsin-2-API concentration. Ⅺ, Stage I; ⅜, Stage II; ‚, Stage lack of trypsinogen-1 and trypsinogen-2 expression predicted III; ϩ, Stage IV.

for adverse outcome. Trypsin-2-API was related to grade, but it 8. Kohn EC, Liotta LA. Molecular insights into cancer invasion: strat- was independent of stage and histological type of the tumor, and egies for prevention and intervention. Cancer Res 1995;55:1856–62. it did not correlate with tissue expression of TATI. 9. Garzetti GG, Ciavattini A, Lucarini G, et al. Tissue and serum The serum concentration of the inactive proenzyme tryp- metalloproteinase (MMP-2) expression in advanced ovarian serous cys- toadenocarcinomas: clinical and prognostic implications. Anticancer sinogen-2 did not predict survival, although it was significantly Res 1995;15:2799–804. greater in patients with high-stage than in those with low-stage 10. Konecny G, Untch M, Pihan A, et al. Association of urokinase-type disease. Elevated serum levels of trypsinogen-2 and trypsin-2- plasminogen activator and its inhibitor with disease progression and API have earlier been observed in a majority of patients with prognosis in ovarian cancer. Clin Cancer Res 2001;7:1743–9. cholangiocarcinoma and pancreatic cancer (45). However, in 11. Chambers SK, Ivins CM, Carcangiu ML. Plasminogen activator these cancers, the elevation may be caused by release of pan- inhibitor-1 is an independent poor prognostic factor for survival in advanced stage epithelial ovarian cancer patients. Int J Cancer 1998;79: creatic trypsin. This is not a likely cause in ovarian cancer. Thus, 449–54. our findings indicate that ovarian cancer secretes trypsin and 12. Koivunen E, Ristima¨ki A, Itkonen O, Osman S, Vuento M, Stenman that this protease contributes to aggressive tumor growth. U-H. Tumor-associated trypsin participates in cancer cell-mediated deg- Trypsin-1 and trypsin-2 are thought to be involved in the radation of extracellular matrix. Cancer Res 1991;51:2107–12. spread of ovarian cancer by degrading extracellular matrix (12) 13. Sorsa T, Salo T, Koivunen E, et al. Activation of type IV procol- and by activating other proteases associated with cancer inva- lagenases by human tumor-associated trypsin-2. J Biol Chem 1997;272: 21067–74. sion, e.g., urokinase-type plasminogen activator and MMPs (13, 14. Lukkonen A, Sorsa T, Salo T, et al. Down-regulation of trypsino- 15, 46). Trypsinogen (17, 18) and TATI (27, 47) are often gen-2 expression by chemically modified tetracyclines: association with expressed in ovarian cancer, and immunohistochemical expres- reduced cancer cell migration. Int J Cancer 2000;86:577–81. sion of trypsinogen-1 has been reported to be more frequent in 15. Paju A, Sorsa T, Tervahartiala T, et al. The levels of trypsinogen malignant than in benign or borderline tumors (19). Trypsino- isoenzymes in ovarian tumour cyst fluids are associated with promatrix gen-1, trypsinogen-2, and TATI occur at high concentrations in metalloproteinase-9 but not promatrix metalloproteinase-2 activation. Br J Cancer 2001;84:1363–71. cyst fluid produced by ovarian tumors (15, 17). These, as well 16. Moilanen M, Sorsa T, Stenman M, et al. Tumor-associated trypsin- as complexes of API with trypsin-1 and trypsin-2, which reflect ogen-2 (trypsinogen-2) activates procollagenases (MMP-1, -8, -13) and trypsin activation, occur at higher concentrations in cyst fluid stromelysin-1 (MMP-13) and degrades type I collagen. Biochemistry from malignant than from benign ovarian tumors (15, 17), and 2003;42:5414–20. this is associated with MMP-9 activation (15). 17. Koivunen E, Itkonen O, Halila H, Stenman U-H. Cyst fluid of The strong prognostic value of TATI expression in ovarian ovarian cancer patients contains high concentrations of trypsinogen-2. Cancer Res 1990;50:2375–8. cancer tissue appears to explain our earlier findings showing 18. Hirahara F, Miyagi Y, Miyagi E, et al. Trypsinogen expression in that serum TATI predicts adverse survival in epithelial ovarian human ovarian carcinomas. Int J Cancer 1995;63:176–81. cancer (3, 30, 31). Serum TATI correlated with prognosis in 19. Hirahara F, Miyagi E, Nagashima Y, et al. Differential expression advanced ovarian cancer also in this study, but an elevated of trypsin in human ovarian carcinomas and low-malignant-potential serum concentration of trypsin-2-API was a prognostic factor tumors. Gynecol Oncol 1998;68:162–5. both among all patients and those with advanced disease. Thus, 20. Ohta T, Terada T, Nagakawa T, et al. Pancreatic trypsinogen and cathepsin B in human pancreatic carcinomas and associated metastatic trypsin-2-API is a new prognostic serum marker, which together lesions. Br J Cancer 1994;69:152–6. with TATI may be useful in the selection of therapy and strat- 21. Fujimura T, Ohta T, Kitagawa H, et al. Trypsinogen expression and ification of patients for clinical studies. early detection for peritoneal dissemination in gastric cancer. J Surg Oncol 1998;69:71–5. 22. Ichikawa Y, Koshikawa N, Hasegawa S, et al. Marked increase of REFERENCES trypsin(ogen) in serum of linitis plastica (gastric cancer, borrmann 4) 1. Friedlander ML. Prognostic factors in ovarian cancer. Semin Oncol patients. Clin Cancer Res 2000;6:1385–8. 1998;25:305–14. 23. Oyama K, Ohta T, Nishimura GI, et al. Trypsinogen expression in 2. Bast RC Jr, Klug TL, St. John E, et al. A radioimmunoassay using a colorectal cancers. Int J Mol Med 2000;6:543–8. monoclonal antibody to monitor the course of epithelial ovarian cancer 24. Williams SJ, Gotley DC, Antalis TM. Human trypsinogen in colo- N Engl J Med 1983;309:883–7. rectal cancer. Int J Cancer 2001;93:67–73. 3. Vartiainen J, Lehtovirta P, Finne P, Stenman, U-H, Alfthan H. 25. Terada T, Ohta T, Minato H, Nakanuma Y. Expression of pancre- Preoperative serum concentration of hCGbeta as a prognostic factor in atic trypsinogen/trypsin and cathepsin B in human cholangiocarcinomas ovarian cancer. Int J Cancer 2001;95:313–6. and hepatocellular carcinomas. Hum Pathol 1995;26:746–52. 4. Frias AE Jr, Li H, Keeney GL, Podratz KC, Woodruff TK. Preop- 26. Yamamoto H, Iku S, Itoh F, Tang X, Hosokawa M, Imai K. erative serum level of inhibin A is an independent prognostic factor for Association of trypsin expression with recurrence and poor prognosis in the survival of postmenopausal women with epithelial ovarian carci- human esophageal squamous cell carcinoma. Cancer (Phila.) 2001;91: noma. Cancer (Phila.) 1999;85:465–71. 1324–31. 5. Tempfer C, Obermair A, Hefler L, Haeusler G, Gitsch G, Kainz C. 27. Stenman U-H, Huhtala M-L, Koistinen R, Seppa¨la¨ M. Immuno- Vascular endothelial growth factor serum concentrations in ovarian chemical demonstration of an ovarian cancer-associated urinary peptide. cancer. Obstet Gynecol 1998;92:360–3. Int J Cancer 1982;30:53–7. 6. Diamandis EP, Scorilas A, Fracchioli S, et al. Human kallikrein 6 28. Huhtala M-L, Pesonen K, Kalkkinen N, Stenman U-H. Purification (hK6): a new potential serum biomarker for diagnosis and prognosis of and characterization of a tumor-associated trypsin inhibitor from the ovarian carcinoma. J Clin Oncol 2003;21:1035–43. urine of a patient with ovarian cancer. J Biol Chem 1982;257:13713–6. 7. Luo L-Y, Katsaros D, Scorilas A, et al. Prognostic value of human 29. Stenman U-H, Koivunen E, Itkonen O. Biology and function of kallikrein 10 expression in epithelial ovarian carcinoma. Clin Cancer tumor-associated trypsin inhibitor, TATI. Scand J Clin Lab Investig Res 2001;7:2372–9. (Suppl) 1991;207:5–9.

30. Venesmaa P, Lehtovirta P, Stenman U-H, Leminen A, Forss M, 39. Yamamoto T, Nakamura Y, Nishide J, et al. Molecular cloning and Ylikorkala O. Tumour-associated trypsin inhibitor (TATI): comparison nucleotide sequence of human pancreatic secretory trypsin inhibitor with CA125 as a preoperative prognostic indicator in advanced ovarian (PSTI) cDNA. Biochem Biophys Res Commun 1985;132:605–12. cancer. Br J Cancer 1994;70:1188–90. 40. Paju A, Bjartell A, Zhang W-M, et al. Expression and character- 31. Venesmaa P, Stenman U-H, Forss M, et al. Pre-operative serum ization of trypsinogen produced in the male genital tract. Am J Pathol level of tumour-associated trypsin inhibitor and residual tumour size as 2000;157:2011–21. prognostic indicators in stage III epithelial ovarian cancer. Br J Obstet 41. Grøndahl-Hansen J, Christensen IJ, Rosenquist C, et al. High levels Gynaecol 1998;105:508–11. of urokinase-type plasminogen activator and its inhibitor PAI-1 in 32. Paju A, Jacobsen J, Rasmuson T, Stenman U-H, Ljungberg B. cytosolic extracts of breast carcinomas are associated with poor prog- Tumor associated trypsin inhibitor as a prognostic factor in renal cell nosis. Cancer Res 1993;53:2513–21. carcinoma. J Urol 2001;165:959–62. 42. Pedersen H, Grøndahl-Hansen J, Francis D, et al. Urokinase and 33. Petterson F, Coppleson MWC, Ludwig H, Shepherd J. Annual plasminogen activator inhibitor type 1 in pulmonary adenocarcinoma. report on the results of treatment in gynecological cancer, 20th ed. Cancer Res 1994;54:120–3. Stockholm: Fe´de`ration Internationale d’Obsterique et Gyne´cologie; 43. Nielsen HJ, Pappot H, Christensen IJ, et al. Association between 1988. plasma concentrations of plasminogen activator inhibitor-1 and survival 34. Itkonen O, Koivunen E, Hurme M, Alfthan H, Schro¨der T, Stenman in patients with colorectal cancer. Br Med J 1998;316:829–30. U-H. Time-resolved immunofluorometric assays for trypsinogen-1 and 44. Zhang W-M, Finne P, Leinonen J, Salo J, Stenman U-H. Determi- 2 in serum reveal preferential elevation of trypsinogen-2 in pancreatitis. nation of prostate-specific antigen complexed to alpha(2)-macroglobulin J Lab Clin Med 1990;115:712–8. in serum increases the specificity of free to total PSA for prostate cancer. 35. Osman S, Turpeinen U, Itkonen O, Stenman U-H. Optimization of Urology 2000;56:267–72. a time-resolved immunofluorometric assay for tumor-associated trypsin 45. Hedstro¨m J, Haglund C, Haapiainen R, Stenman U-H. Serum tryp- inhibitor (TATI) using the streptavidin-biotin system. J Immunol Met ␣ sinogen-2 and trypsin-2- 1-antitrypsin complex in malignant and benign 1993;161:97–106. digestive-tract diseases. Preferential elevation in patients with cholan- 36. Halila H, Lehtovirta P, Stenman U-H. Tumour-associated trypsin giocarcinomas. Int J Cancer 1996;66:326–31. inhibitor (TATI) in ovarian cancer. Br J Cancer 1988;57:304–7. 46. Koivunen E, Huhtala ML, Stenman U-H. Human ovarian tumor- 37. Hedstro¨m J, Leinonen J, Sainio V, Stenman U-H. Time-resolved associated trypsin. Its purification and characterization from mucinous immunofluorometric assay of trypsin-2 complexed with ␣-1-antitrypsin cyst fluid and identification as an activator of pro-urokinase. J Biol in serum. Clin Chem 1994;40:1761–5. Chem 1989;264:14095–9. 38. Emi M, Nakamura Y, Ogawa M, et al. Cloning, characterization and 47. Ueda G, Shimizu C, Tanaka Y, et al. Immunohistochemical dem- nucleotide sequences of two cDNAs encoding human pancreatic trypsi- onstration of pancreatic secretory trypsin inhibitor in gynecologic tu- nogens. Gene (Amst.) 1986;41:305–10. mors. Gynecol Oncol 1989;32:37–40.

Downloaded from clincancerres.aacrjournals.org on October 2, 2021. © 2004 American Association for Cancer Research. Expression of Trypsinogen-1, Trypsinogen-2, and Tumor-Associated Trypsin Inhibitor in Ovarian Cancer: Prognostic Study on Tissue and Serum

Annukka Paju, Juhani Vartiainen, Caj Haglund, et al.

Clin Cancer Res 2004;10:4761-4768.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/10/14/4761

Cited articles This article cites 44 articles, 14 of which you can access for free at: http://clincancerres.aacrjournals.org/content/10/14/4761.full#ref-list-1

Citing articles This article has been cited by 5 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/10/14/4761.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/10/14/4761. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.