SEARCH

SEARCH BY CITATION

Keywords:

  • biomarker;
  • bladder cancer;
  • plasminogen activator inhibitor type 1;
  • prognosis;
  • survival

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

BACKGROUND:

Recent studies have demonstrated a poor prognosis for patients who have altered expression of plasminogen activator inhibitor type 1 (PAI-1) in several cancer types. The objective of the current study was to investigate the prognostic impact of PAI-1 on patients with transitional cell carcinoma (TCC) of the urinary bladder.

METHODS:

PAI-1 expression was quantified using real-time polymerase chain reaction in 91 TCCs and in 6 normal tissue specimens. PAI-1 concentrations were analyzed by enzyme-linked immunoadsorbent assay in plasma from 104 patients and 10 controls and in urine from 244 patients and 74 controls. PAI-1 expression was evaluated immunohistochemically in paraffin-embedded tissues (94 tumor samples and 10 adjacent normal tissue samples). The results were analyzed in relation to clinical features and follow-up.

RESULTS:

Significantly higher PAI-1 levels were detected in tissue and plasma samples, but not in urine, from patients with bladder cancer compared with controls (P = .001 and P = .008, respectively). Elevated gene expression and plasma protein concentrations were independent of tumor stage and grade. Immunostaining revealed a subgroup of patients with single tumor cells that strongly expressed PAI-1. These patients' survival was significantly shorter, and their clinical presentation was correlated significantly with lymph node-positive disease.

CONCLUSIONS:

PAI-1 gene expression in tissues and plasma protein levels were elevated in patients with TCC compared with controls. PAI-1 gene or protein expression was not associated with the clinical characteristics of bladder TCC. Although the assessment of PAI-1 expression in plasma or urine may not serve as an independent predictor of prognosis in patients with TCC, the immunohistochemical detection of single PAI-1–positive cells may serve as a predictor of survival and a possible indicator of metastasis. Cancer 2010. © 2010 American Cancer Society.

Cancer invasion and metastasis develop through a sequence of processes that involve the loss of cell-cell and cell-matrix adhesions, proteolysis, and the induction of angiogenesis.1, 2 Different protease systems are involved in these processes, especially in invasion and the development of metastases. One of these systems is the plasminogen activator (PA) system, in which the serine protease urokinase-type PA (uPA) is central.3-5 Another PA is the tissue-type PA (tPA). Unlike uPA, tPA has a strong affinity for fibrin and is considered to be involved primarily in thrombolysis. The activity of PA can be neutralized by 2 main, specific PA inhibitors (PAIs): PAI type 1 (PAI-1) and PAI-2.5 Results from different human cancers have demonstrated that the levels of components of the PA system in tumors are significantly higher than the levels in corresponding normal tissue.6, 7 More important, high levels of uPA and PAI-1 in malignant tumors are correlated with a poor prognosis. For example, uPA and PAI-1 are among the strongest prognostic factors described to date for breast cancer.6

PAI-1 is a multifunctional protein with various tumor-promoting characteristics. Mechanistically, PAI-1 mediates cell migration,8 stimulates angiogenesis,9 and modulates cell adhesion.10 These findings suggest that PAI-1 contributes directly to the pathology of cancer. The effects of PAI-1 on the development of cancer probably are more important than its inhibiting effect on the fibrinolytic system.11-13

Bladder cancer (BC) is a very common malignant tumor of the genitourinary tract. BC, clinically diagnosed as a superficial tumor, has a relatively indolent nature and low malignant potential. However, muscular invasion of the tumor is associated with a high risk of subsequent metastasis and poor survival.14, 15 Current therapeutic management is based on conventional prognostic factors (ie, tumor stage and grade). Identification of new molecular markers potentially could lead to improvements in prognosis and the development of novel therapeutic approaches using targeted treatment modalities. To date, little is known about the prognostic impact of PAI-1 in BC. In this study, we investigated distinct the profiles of PAI-1 expression on the gene and protein levels and elucidated their impact on the survival of patients with transitional cell cancer (TCC) of the bladder. The current results identified a subgroup of patients with BC at high risk of a poor prognosis.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

Clinical Samples

In total, 633 individuals (533 patients with TCC and 100 controls) divided in 4 cohorts were included in this study; 1) frozen tissue samples from 91 patients with BC and 6 controls were analyzed by quantitative real-time polymerase chain reaction (PCR); 2) plasma samples from 104 patients with BC and 10 healthy individuals and 3) urine samples from 244 patients with BC and 74 controls were analyzed by enzyme-linked immunosorbent assay (ELISA); and 4) paraffin-embedded tissue blocks from 94 patients with BC and 10 controls were used for immunohistochemistry (Tables 1 and 2). We obtained tumor and plasma samples from patients who underwent surgical treatment for BC in the Department of Urology of the University Hospital of Essen between 1992 and 1999. Urine samples were collected between 2005 and 2007. Criteria for enrollment were a histopathologic diagnosis of bladder TCC, no history of other tumor, no chemotherapy before surgery, the availability of sufficient tumor sample, and the potential to be followed. The Ethics Committee of the University Hospital of Essen approved the study protocol. All tumors were reclassified according to the World Health Organization classification of urothelial neoplasms for the year 2004.16 Tissue sections from each biopsy were stained with hematoxylin and eosin and were reviewed by a pathologist to define the tumor type and stage. Only biopsies that contained ≥50% tumor cells were selected for analysis. Normal bladder epithelia was used as a control in PCR analyses that originated from an open enucleation in patients with benign prostatic hyperplasia. Insufficient PCR analyses with a high standard deviation (>0.5) between parallel measurements were repeated.

Table 1. Patient Characteristics: Plasminogen Activator Inhibitor Type 1 in Different Patient Biomaterials
CharacteristicPAI-1 Gene ExpressionPAI-1 Plasma Concentration, ng/mLPAI-1 Urine Concentration, ng/mL
No.Median (Range)PNo.Median (Range)PNo.Median (Range)P
  • PAI-1 indicates plasminogen activator inhibitor type 1; TCC, transitional cell carcinoma.

  • a

    The median patient age was 66.1 years (range, 41-95 years).

Age, ya         
 ≤65410.69 (0.02-18.76).648464.29 (1.66-74.26).348980.14 (0-33.86).076
 >65510.76 (0.09-76.54) 585.72 (1.51-86.39) 1460.30 (0-26.00) 
Sex         
 Women230.88 (0.09-63.43).589195.97 (1.66-68.23).407940.05 (0-21.68).037
 Men680.74 (0.02-76.54) 854.42 (1.51-86.39) 1500.29 (0-33.86) 
Tumor classification         
 Control60.06 (0.03-0.28).001102.40 (0.84-5.12).059740.19 (0-1.35).271
 Ta300.54 (0.07-13.13).926254.19 (1.61-86.39).873790.04 (0-14.40).007
 T1190.45 (0.04-5.4).880203.87 (1.70-31.17).150920.40 (0-23.43).954
 T2130.58 (0.1-1.73).080377.53 (1.67-74.26).143600.36 (0-26.00).420
 T3201.04 (0.02-76.54).945153.35 (1.51-60.82).50370.00 (0-14.32).534
 T490.9 (0.03-18.76) 75.47 (2.20-18.03) 61.08 (0-33.86) 
 Noninvasive490.45 (0.04-13.13).164454.07 (1.61-86.39).3551710.15 (0-23.43).266
 Invasive420.82 (0.02-76.54) 595.47 (1.51-74.26) 730.29 (0-33.86) 
Grade         
 1200.56 (0.07-13.13).657213.06 (1.66-86.39).654370.00 (0-10.73).133
 2310.45 (0.04-9.24).065354.07 (1.51-68.23).1771330.14 (0-23.43).003
 3400.93 (0.02-76.54) 486.48 (1.66-74.26) 740.75 (0-33.86) 
 Low grade510.45 (0.04-13.13).065563.90 (1.51-86.39).0861700.08 (0-23.43).001
 High grade400.94 (0.02-76.54) 486.48 (1.66-74.26) 740.75 (0-33.86) 
Tumor occurrence         
 Primary570.69 (76.54-0.04).421514.65 (1.51-74.26).904730.34 (0-33.86).318
 Recurrent340.77 (0.02-13.13) 534.51 (1.61-86.39) 1090.13 (0-14.40) 
Smoking         
 No340.67 (0.02-18.76).614198.11 (1.94-68.23).129270.27 (0-5.50).179
 Yes300.57 (0.07-11.73) 224.58 (1.96-3.87) 350.38 (0-33.86) 
 Unknown17  63  182  
Control60.06 (0.03-0.28).001102.40 (0.84-5.12).008740.18 (0-1.11).232
TCC tumor910.74 (0.02-76.54) 1044.46 (1.51-86.39) 2440.19 (0-33.86) 
Table 2. Patient Characteristics: Immunostaining for Plasminogen Activator Inhibitor Type 1
Immunostaining for PAI-1No. of Patients (%)P
All PatientsPatients With Single PAI-1- Positive CellsPatients With No Single PAI-1- Positive Cells
  1. PAI-1 indicates plasminogen activator inhibitor type 1; TCC, transitional cell carcinoma.

All patients94 (100)17 (18)77 (82) 
Age, y   .406
 ≤6534 (36)6 (35)28 (36) 
 >6560 (64)11 (65)49 (64) 
Sex   .350
 Women23 (24)6 (35)17 (22) 
 Men71 (76)11 (65)60 (78) 
Tumor classification   .002
 Ta31 (33)2 (12)29 (38) 
 T124 (26)1 (6)23 (30) 
 T227 (29)10 (59)17 (22) 
 T38 (9)3 (18)5 (6) 
 T44 (4)1 (6)3 (4) 
 Noninvasive: Ta-T155 (59)3 (18)52 (68)<.001
 Invasive: T2-T439 (41)14 (82)25 (32) 
Grade   .018
 118 (19)0 (0)18 (23) 
 247 (50)8 (47)39 (51) 
 329 (31)9 (53)20 (26) 
 Low: Grade 1 and 265 (69)8 (47)57 (74).139
 High: Grade 329 (31)9 (53)20 (26) 
Lymph node status   <.001
 Positive7 (7)6 (35)1 (1) 
 Negative56 (60)8 (47)48 (62) 
 Unknown31 (33)3 (18)28 (36) 
TCC941777<.001
Control10010 

RNA Isolation and Combinational DNA Synthesis

After homogenization in QIAzol reagent (Qiagen, Hilden, Germany), RNA isolation was done according to the manufacturer's instructions. The isolated RNA was further purified using the RNeasy Mini kit (Qiagen). DNA was digested with RNase-free DNase Set (Qiagen). Total RNA was quantified using an UV spectrophotometer (ND-1000; Peglab Biotechnology GmbH, Erlangen, Germany). The quality and integrity of samples were assessed on a 1.5% agarose gel. RNA was reverse transcribed in a final volume of 20 μL containing 200 ng RNA, reverse transcription buffer, 0.5 mmol/L deoxynucleotriphophates, 1.8 μmol/L oligo(dT), 10 U RNase inhibitor, and 40 U Omniscript RTase. Combinational DNA (cDNA) synthesis was carried out at 37°C for 60 minutes.

TaqMan 2-Step Reverse Transcriptase-PCR Assay

Quantitative real-time PCR was done using 48-well plates on a StepOne real-time PCR system (Applied Biosystems, Hamburg, Germany). To provide high reproducibility, we used the predeveloped TaqMan Gene Expression Assay (Applied Biosystems). The assay identification was as follows: PAI-1 (Hs01126602_m1), and the TATA box-binding protein 4333769. The expression levels were related to Universal Human Reference RNA (Stratagene, La Jolla, Calif). We used the TATA box-binding protein to normalize target gene expression. PCRs contained 2.5 μL cDNA, 1×TaqMan Gene Expression Assay, and 1×TaqMan Universal PCR Master Mix (Applied Biosystems) in a 25-μL volume. The thermal cycling conditions were as follows: 95°C for 10 minutes, followed by 40 cycles at 95°C for 15 seconds, and 60°C for 1 minute. Each sample was tested in duplicate. The ΔΔCt approximation method was used to calculate gene expression values.17

Measurement of Plasma und Urine PAI-1 Levels

Levels of human serpinE1/PAI-1 were quantified by using a sandwich ELISA with the Duoset ELISA kit from R&D Systems (Wiesbaden, Germany) according to the manufacturer's instructions. All samples were examined in triplicate. Mean values were used for statistical analysis.

Immunostaining

Immunohistochemical staining was done by minor modifications of the procedure described previously.18 Sections were cut to a thickness of 2 to 4 μm and mounted on protein-coated glass slides. After dewaxing in xylene and rehydration in a series of alcohols, endogenous peroxidase activity was blocked with 5% hydrogen peroxidase followed by incubation with the primary antibody against PAI-1 (dilution, 1:200; Santa Cruz Biotechnology, Santa Cruz, Calif) for 30 minutes at room temperature using an automated autostainer (Dako-Cytomation, Glostrup, Denmark). Then, a second incubation with Zytomed POLHRP-100 (brown) or POLAP-100 (red) as well as visualization according to the manufacturer's instructions (Zytomed Systems, Berlin, Germany) was performed. The specificity of the immunoreactions was checked by omission of the primary antibody. Breast cancer served as positive control.

For double immunohistochemistry, the same PAI-1 antibody and the monoclonal cytokeratin (CK) antibody CK-MNF (dilution 1:1000; Dako-Cytomation) were used. Antigen retrieval was carried out at 98°C for 10 minutes in a water bath (target retrieval buffer; DAKO, Glostrup, Denmark). PAI-1/CK-MNF double labeling was carried out in 2 steps. First, PAI-1 labeling was done using an immunohistochemical staining technique based on a horseradish peroxidase (HRP)-labeled polymer conjugated to secondary antibodies (ZytoChemPlus HRP Polymer Kit; Zytomed Systems). Staining was completed by an incubation with 3,3′-diaminobenzidine plus substrate-chromogen (Zytomed Systems), which resulted in a brown-colored precipitate at the antigen site. Second, CK-MNF labeling was done with an alkaline phosphatase-labeled polymer (ZytoChemPlus AP Polymer Kit; Zyomed Systems) and was developed with a Permanent Red chromogenic substrate system (Zytomed Systems). At the end of the procedure, nuclei were counterstained with hematoxylin for 5 minutes.

Microscopic Analysis

Two investigators who were unaware of clinical data independently evaluated PAI-1 staining under a light microscope at ×400 magnification. They graded the staining intensity on a scale from 0 to 3 (with a higher number indicating a higher intensity). Renal and liver control tissues served as internal controls. For the detection of single PAI-1–positive cells (SPPCs) and for discordant cases concerning overall staining intensity, the results were reviewed by both investigators together.

Statistical Analysis

For statistical analysis, the nonparametric, 2-sided Wilcoxon rank-sum test for paired group comparisons was applied. Univariate analysis was done using both the Kaplan-Meier log-rank test and univariate Cox analysis. In immunohistochemistry, the baseline characteristics of the 2 groups of patients (with or without SPPCs) were compared by using chi-square tests. In all tests, P values ≤.05 were considered statistically significant. All statistical analyses were done with the SPSS software package (17.0; SPSS, Inc., Chicago, Ill).

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

Clinical Background

The main characteristics of all 633 patients are provided in Tables 1 and 2. The median follow-up was 41 months, and the maximum follow-up was 196 months. Clinical follow-up data were available for fresh-frozen and paraffin-embedded tumor samples and plasma but not for urine samples.

Comparison of PAI-1 Between Patients and Controls

Tissue messenger RNA expression

PAI-1 messenger RNA (mRNA) expression was measured in 6 samples of normal bladder epithelium and in 91 samples of bladder TCC. We detected significantly higher (12-fold) PAI-1 gene expression in tumor samples compared with samples of normal epithelium (P = .001). No relation between PAI-1 and patient sex, age, smoking consumption, or primary versus recurrent cancer status was observed (Table 1).

Plasma concentration

One hundred four preoperative plasma samples from patients with TCC and from 10 healthy controls were analyzed using the PAI-1 ELISA. PAI-1 protein concentrations were elevated significantly in plasma from the patients with TCC (P = .008). There were no significant differences in PAI-1 concentrations between men and women or between patients with primary cancer versus recurrent cancer. Plasma concentrations did not correlate with age or smoking consumption (Table 1).

Urine concentration

Two hundred forty-four samples of mornings' first urine from patients with TCC and from 74 controls were analyzed using the PAI-1 ELISA. We observed no difference between patients versus controls but observed 6-fold higher PAI-1 concentrations in men compared with women (P = .037). Urinary PAI-1 concentrations were not correlated with age, smoking consumption, or primary versus recurrent cancer but were correlated with hematuria (Table 1).

Relation Between PAI-1 Tissue Gene Expression, Plasma Concentration, Urine Concentration, and Clinicopathologic Parameters

We did not detect any significant correlations between tissue (mRNA) and plasma PAI-1 levels and tumor stage or grade. In urine, we observed 9.5-fold higher concentrations in patients with high-grade BC versus patients with low-grade BC patients but no correlation with tumor stage (Table 1).

Univariate Analysis: Patient Prognosis

In a univariate analysis of clinicopathologic parameters and prognostic endpoints (overall survival, disease-specific survival, and metastasis-free survival), the patients were subdivided into low and high groups according to PAI-1 mRNA gene expression with the median set as the cutoff value. Patient age, sex, and smoking habits did not influence disease-specific survival, metastasis-free survival, or recurrence-free survival. In contrast, tumor stage and grade (low vs high) were strong predictors of reduced disease-specific survival (P < .001 for both) and metastasis-free survival (P < .001 for both). We did not observe any association between high versus low PAI-1 mRNA gene expression and patient prognosis. None of the observed clinicopathologic parameters that we evaluated predicted disease recurrence. Further univariate analysis of plasma and urine results did not any reveal additional information regarding the prognostic endpoints (data not shown).

Immunostaining and SPPCs in BC Tissue

We were not able to establish that, as we assumed in previous investigations, PAI-1 is a predictor of prognosis in patients with TCC; however, we did detect significantly higher PAI-1 concentrations in tissue and plasma from patients with cancer compared with the levels under normal conditions. To elucidate the reason for this difference, we carried out immunohistochemical staining of samples from 94 patients with TCC and 10 controls. In parallel, we also analyzed PAI-1 expression in tissue sections from 20 patients with high and low urinary PAI-1 concentrations (10 patients with high concentrations vs 10 patients with low concentrations) by using immunohistochemistry to investigate the correlation between tissue and urinary PAI-1 concentrations, and we observed no correlation (data not shown). Characteristics of the patients who were analyzed by immunohistochemistry are listed in Table 2. Two independent observers did not identify any relation between the overall PAI-1 staining intensity and parameters like age, sex, and primary versus recurrent cancer, or with clinicopathologic findings, such as disease stage or tumor grade (data not shown). However, in 17 of 94 patients (18%), single tumor cells that strongly expressed PAI-1 were observed (Fig. 1). These cells were localized within the epithelium of superficial (pathologic Ta [pTa]) tumors (Fig. 1A) but mainly were observed within invasive solid tumor cell clusters (Fig. 1B) and in the lumen of bladder lymphatic vessels that contained tumor cell clusters (Fig. 1C). In some tumor areas, PAI-1–positive cells appeared to form polynuclear cells (Fig. 1D). Double immunostaining with antibodies against PAI-1 and cytokeratin 20 confirmed the urothelial origin of these tumor cells. Furthermore, PAI-1 immunostaining was observed in the endothelial cells of some blood vessels, particularly arteries in the surrounding tissue, but such staining also was observed far from the tumor area (Fig. 2).

thumbnail image

Figure 1. These photomicrographs of immunohistochemistry for plasminogen activator inhibitor type 1 (PAI-1)-positive cells reveal PAI-1 staining in bladder cancer tissues. Single cells or small groups of cells (arrows) that were positive for PAI-1 are detectable in (A) in a superficial pathologic Ta tumor, (B) within clusters of invasive tumor cells, and (C) within the lumen of lymphatics (Ly) invaded by tumor cells (TC). (D) Some PAI-1–postive tumor cells display a polynuclear morphology (arrows).

Download figure to PowerPoint

thumbnail image

Figure 2. Plasminogen activator inhibitor type 1 (PAI-1) immunostaining is shown in blood vessels from bladder tissues. (A) Although part of a blood vessel in the tumor-surrounding area demonstrated clear PAI-1 staining in endothelial cells (arrows), (B) another part of a blood vessel (BV) in this area remained negative. (C) In some areas of the invasive front of tumor tissue, single PAI-1–positive cells were noted in the lumen of tumor-associated blood vessels (arrow). (D) Endothelial cells in a portion of arteries far from the tumor area also exhibited PAI-1 staining (arrows).

Download figure to PowerPoint

Focusing our particular interest on these patients, we identified the presence of SPPCs as a strong and significant predictor of disease-specific survival (P = .006) in univariate analysis (Table 3). Kaplan-Meier curves revealed reduced cancer-specific survival when SPPCs were present (Fig. 3, top). SPPCs were not identified in specimens of normal adjacent control tissues, and their presence was correlated with muscle-invasive disease (P < .001). Selecting only patients with muscle-invasive disease for an additional Kaplan-Meier analysis, the presence of SPPCs remained a significant predictor of cancer-specific survival, especially for a short-term prediction time of 12 months (Fig. 3, bottom). Fifty percent of patients with invasive BC who expressed SPPCs died within 12 months. In contrast, the absence of these cells was accompanied by a 12-month survival rate of 80% (Fig. 3, bottom).

Table 3. Cox Regression Univariate Analysis, Immunohistochemistry, Single Plasminogen Activator Inhibitor Type 1-Positive Cells, and Prognostic Endpoints
VariableNo. of PatientsDisease-Specific SurvivalMetastasis-Free Survival
HR95% CIPHR95% CIP
  • HR indicates hazard ratio; CI, confidence interval; IHC, immunohistochemistry.

  • a

    The median patient age was 67.2 years (range, 40-84 years).

Age, ya       
 ≤6534Referent  Referent  
 >65601.2530.619-2.538.5310.5750.228-1.431.241
Sex       
 Women23Referent  Referent  
 Men710.9470.470-1.911.880.9280.355-2.427.879
Stage       
 Noninvasive55Referent  Referent  
 Invasive395.6072.570-12.233<.00126371.008-6.897.048
Grade       
 Low grade65Referent  Referent  
 High grade293.0121.511-6.006.00234371.346-8.780.010
Tumor occurrence       
 Primary68Referent  Referent  
 Recurrence261.0340.695-1.537.8691.2700.683-2.362.451
Smoking       
 No24Referent  Referent  
 Yes200.8240.562-1.342.7241.1020.784-1.366.689
 Unknown50      
Single positive cells (IHC)       
 No77Referent  Referent  
 Yes174.5481.646-12.568.0061.8410.527-6.432.339
thumbnail image

Figure 3. Kaplan-Meier estimates of cancer-specific survival are shown. (Top) Cancer-specific survival is illustrated according to the presence or absence of a single plasminogen activator inhibitor type 1 (PAI-1)-positive (pos) cell on immunohistochemistry. Follow-up was 120 months. All patients (Ta-T4 tumors) were included. (Bottom) Cancer-specific survival is illustrated according to the presence or absence of a single PAI-1–positive cell on immunohistochemistry. Follow-up was 12 months. Only patients with muscle-invasive disease (T2-T4 tumors) were included in this analysis.

Download figure to PowerPoint

Furthermore, the presence of SPPCs was correlated positively (P < .001) with lymph node-positive disease. Six of 7 patients (86%) with positive lymph node status had SPPCs identified, whereas only 8 of 56 patients (14%) with negative lymph node status had these cells identified.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

In this study, we demonstrated significantly elevated PAI-1 concentrations in BC tissues and enhanced PAI-1 concentrations in plasma from patients who had BC compared with controls. The current results reveal a subgroup of patients with BC who are at high risk of a poor prognosis. In this subgroup, we demonstrated for the first time the existence of SPPCs, which were correlated significantly with reduced cancer-specific survival. Furthermore, the detection of SPPCs was attributed to invasive disease and metastatic spread to the lymph nodes.

The degradation of proteins in the extracellular matrix and the basement membrane is necessary for cancer cell invasion and metastasis. One of the enzyme systems involved in the disintegration of the extracellular matrix is the PA system. PA is catalyzed by serine proteases. The major inhibitor of these enzymes is PAI-1. On the basis of the ability of PAI-1 to inhibit proteolysis, high PAI-1 levels in a tumor would be expected to inhibit cancer cell growth and dissemination in patients and, thus, would be related to a more favorable prognosis. However, clinical studies to date have demonstrated consistently that PAI-1 is strongly related to a poor outcome in several types of cancer.4, 6, 7, 19-21 The reason for this apparent discrepancy still is not understood fully. PAI-1 seems to be capable of various tumor-promoting characteristics, which reportedly are more important than its inhibiting effects on the PA system.12, 13 Consistent with those reports, our data indicate that plasma PAI-1 levels are increased in BC and, more important, that the presence of SPPCs within the BC tissue is associated significantly with a poor prognosis.

To our knowledge, the only study investigating plasma levels in urinary tract carcinoma dates from 1994.19 In that study, Bashar et al demonstrated higher PAI-1 concentrations in plasma from patients with versus without metastatic disease and in preoperative versus postoperative patients after the removal of cancer tissue. Our results confirmed the general aspect of elevated PAI-1 levels in plasma from patients with cancer; however, we did not observe an association between plasma PAI-1 levels and lymph node status. Our plasma analysis did not confirm the predictive value for BC, unlike the results reported by other groups, who which observed an association between plasma PAI-1 levels and survival or disease recurrence in other tumor entities, such as nonsmall cell lung carcinoma20 and breast cancer.21

Urinary PAI-1 has been correlated with diabetic nephropathy.22 In a further step, we also determined urinary PAI-1 levels in patients with BC to explore whether a similar relation may exist between urinary PAI-1 levels and BC. Although our findings revealed significantly higher concentrations (10-fold) of PAI-1 in urine samples from patients with high-grade BC versus low-grade BC, we also observed a significant correlation between urinary PAI-1 levels and hematuria. Because platelets have been considered the main source of released PAI-1,23 we hypothesized that the high levels of PAI-1 in urine samples from patients with invasive BC may be caused by the release of PAI-1 from the platelets and its secretion in the urine by the process of hematuria rather than by tumor cells themselves. Furthermore, we observed that high-grade BC was more likely to be accompanied by hematuria, and this may explain the elevated PAI-1 levels in our patients with high-grade disease. Supporting this interpretation, immunohistochemical analyses on paraffin sections of tumor tissues from 10 patients with “high” versus 10 patients with “low” urinary PAI-1 concentrations did not reveal any differences in the expression patterns. This finding also may explain the inconsistencies between the urinary and plasma PAI-1 concentrations. For these reasons, the measurement of urinary PAI-1 fails to provide additional information regarding diagnosis or prognosis in patients with BC.

In tissue expression studies, Sternlicht et al observed that high PAI-1 mRNA levels predicted shorter overall survival in 2 independent breast cancer datasets.24 PAI-1 antigen levels determined by ELISA in primary tumor tissue extracts from patients with ovarian cancer also has been identified as a statistically independent prognostic factor.25 To determine PAI-1 tissue expression levels, recent investigations almost exclusively have used ELISA on tumor tissue extracts along with immunohistochemistry. We investigated gene expression levels using real-time PCR. In the current study, we observed 12-fold elevated mRNA levels in tumors compared with normal tissues. This is in accordance with the recent findings of Span et al, who used an ELISA to demonstrate higher PAI-1 levels in renal cell carcinoma and BC compared with the levels in normal tissues.13 In that study, there were no data on the prognostic impact of the observations. We compared PAI-1 gene expression levels with clinical follow-up data. In addition to significant differences between tumors and normal tissues, we observed 2-fold elevated PAI-1 expression in invasive bladder TCC compared with noninvasive bladder TCC. These findings underline the obvious involvement of PAI-1 in the development (and possibly in the invasion) of BC. Conversely, we did not observe any correlation between PAI-1 gene expression levels and disease-specific survival, recurrence-free survival, or metastasis-free survival as reported previously (mainly in the gynecologic field). Discrepancies in the results from different groups concerning the prognostic impact of PAI-1 in tumor tissue may be explained by recent investigations: It is noteworthy that the in vivo experiments of Chen et al demonstrated inhibiting effects on tumor invasion and progression after the administration of PAI-1 in an orthotopic rat bladder tumor model.26 In contrast, in vitro data of Kwaan et al suggested that PAI-1 has tumor-promoting properties. In their study, the transfection of recombinant PAI-1 into the human prostate cell line PC-3 protected the cells against chemotherapy-induced apoptosis.11 Obviously, the predominant impact of PAI-1 on the biologic behavior of tumors differs under certain conditions. Interpreting these conflicting results, we hypothesized that it is crucial whether PAI-1 is expressed by tumor cells or by the host.

This assumption led us to determine the location of PAI-1 expression in bladder TCC by using immunohistochemistry. PAI-1 expression was located in the cytoplasm of tumor cells of epithelial origin. The presence of single or small groups of cells was detectable mainly within invasive tumor cell clusters or in the lumen of lymphatics that were invaded by tumor cells (Fig. 1). It is interesting to note that other investigators studying breast cancer have suggested that PAI-1 serves to protect the tumor tissue itself against the proteolytic degradation that the tumor imposes on the surrounding normal tissue. In contrast, expression of PAI-1 by the hosting stroma may provide protection from the proteolytic activities of the invading tumor.27 Our visual and statistical observations underline the hypothesis that PAI-1 is involved in cancer progression. The existence of single cells that expressed PAI-1 was correlated significantly with muscle and lymph vessel invasion. To better characterize these cells, we performed additional double immunostaining with PAI-1 and CK-20. CK-20 positivity confirmed the epithelial origin of these cells and also revealed the partial loss of CK-20 expression in deeper invading, PAI-1–expressing tumor cells. This loss could be interpreted as an epithelial-to-mesenchymal transition (EMT), a process in which tumor cells lose their epithelial markers and start to express mesenchymal markers. EMT is a feature of aggressive tumors and is characterized by increased motility and invasion, which have been observed in several human cancers.28

Our Kaplan-Meier analysis demonstrated that PAI-1 expression may be a feature of very aggressive tumor cells and confirmed the findings of several in vitro studies that analyzed the potential tumor-promoting properties of PAI-1.9-12 Because of a stage-dependent effect (almost all patients who had specimens with SPPCs had T2-T4 tumors), multivariate analysis could not verify this factor as independent predictor of prognosis. Restricting the survival analysis only to patients with muscle-invasive disease, we were able to select a subgroup of patients at high risk of poor survival (Fig. 3, bottom). These patients potentially could benefit from targeted therapy in the future. New inhibitory agents against PAI-1 recently have produced encouraging (but not yet clinically verified) results in experimental tumors.23

Along with its potential as a predictive marker for BC, it has been demonstrated that PAI-1 plays a crucial role in vessel formation.9, 29 PAI-1 deficiency has been associated with reduced vessel sprouting,30 and PAI-1 null mice exhibited reduced tumor growth mainly because of decreased tumor vascularization29, 31; whereas, at supraphysiologic concentrations, PAI-1 suppressed both the vascularization and the growth of tumors. The high PAI-1 expression levels in SPPCs within BC tissue may build PAI-1 levels up to concentrations that promote angiogenesis (eg, at the invasive front of BC tissue, where the highest number of SPPCs were located). Further studies that include uPA and tPA will be needed to better assess the role of PAI-1 and the PA system in BC growth, invasion, and vascularization. Larger patient numbers and prospective studies and the integration of complementary approaches will enable us to enhance its use as a prognostic and predictive tool or as a therapeutic target in the future.

In conclusion, previously published data about the prognostic impact of PAI-1 in other types of cancer, independent of established risk factors like pathologic tumor classification or lymph node status, allowed us to hypothesize that PAI-1 may be a strong candidate as a marker for the identification of aggressive urothelial tumors regardless of clinicopathologic stage. To our knowledge, the current study is the first to date that systematically demonstrates the prognostic impact of PAI-1 gene and protein expression in tissue, plasma, and urine from patients with BC. Our analysis demonstrated higher PAI-1 levels in tissue and plasma from patients with BC compared with healthy controls. Although our results suggest that the assessment of PAI-1 gene expression and the measurement of urine and plasma PAI-1 levels apparently are not useful for predicting the prognosis of patients with TCC, the immunohistochemical detection of SPPCs was associated significantly with poor survival and lymph node metastasis. Furthermore, it can be postulated that PAI-1 produced and secreted from SPPCs at the invasive front of BC can support new vessel formation and, thus, tumor vascularization, as indicated by previous studies. The current study provides a basis for further analyses to explore both the (patho)physiologic role of PAI-1 and other members of the PA system, such as uPA and tPA, in BC and their possible relation to vascularization, progression, and prognosis in this tumor type.

Acknowledgements

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

We thank Andre Scherag, MD (Department of Biostatistics) for his statistical assistance.

CONFLICT OF INTEREST DISCLOSURES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES

Supported by a grant from the German National Federal Ministry of Education&Research (grant 0313659B).

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. Acknowledgements
  7. CONFLICT OF INTEREST DISCLOSURES
  8. REFERENCES
  • 1
    Gontero P, Banisadr S, Frea B, Brausi M. Metastasis markers in bladder cancer: a review of the literature and clinical considerations. Eur Urol. 2004; 46: 296-311.
  • 2
    Becker M, Tilki D, Szarvas T, Rubben H, Ergun S. Urine-based markers of angiogenesis in bladder cancer. Urologe A. 2009; 48: 609-614.
  • 3
    Schmitt M, Harbeck N, Thomssen C, et al. Clinical impact of the plasminogen activation system in tumor invasion and metastasis: prognostic relevance and target for therapy. Thromb Haemost. 1997; 78: 285-296.
  • 4
    Duffy MJ, Maguire TM, McDermott EW, O'Higgins N. Urokinase plasminogen activator: a prognostic marker in multiple types of cancer. J Surg Oncol. 1999; 71: 130-135.
  • 5
    Rijken DC. Plasminogen activators and plasminogen activator inhibitors: biochemical aspects. Baillieres Clin Haematol. 1995; 8: 291-312.
  • 6
    Harbeck N, Kates RE, Gauger K, et al. Urokinase-type plasminogen activator (uPA) and its inhibitor PAI-I: novel tumor-derived factors with a high prognostic and predictive impact in breast cancer. Thromb Haemost. 2004; 91: 450-456.
  • 7
    Berger DH. Plasmin/plasminogen system in colorectal cancer. World J Surg. 2002; 26: 767-771.
  • 8
    Stahl A, Mueller BM. Melanoma cell migration on vitronectin: regulation by components of the plasminogen activation system. Int J Cancer. 1997; 71: 116-122.
  • 9
    Bajou K, Noel A, Gerard RD, et al. Absence of host plasminogen activator inhibitor 1 prevents cancer invasion and vascularization. Nat Med. 1998; 4: 923-928.
  • 10
    Palmieri D, Lee JW, Juliano RL, Church FC. Plasminogen activator inhibitor-1 and -3 increase cell adhesion and motility of MDA-MB-435 B breast cancer cells. J Biol Chem. 2002; 277: 40950-40957.
  • 11
    Kwaan HC, Wang J, Svoboda K, Declerck PJ. Plasminogen activator inhibitor 1 may promote tumour growth through inhibition of apoptosis. Br J Cancer. 2000; 82: 1702-1708.
  • 12
    Lademann UA, Romer MU. Regulation of programmed cell death by plasminogen activator inhibitor type 1 (PAI-1). Thromb Haemost. 2008; 100: 1041-1046.
  • 13
    Span PN, Witjes JA, Grebenchtchikov N, et al. Components of the plasminogen activator system and their complexes in renal cell and bladder cancer: comparison between normal and matched cancerous tissues. BJU Int. 2008; 102: 177-182.
  • 14
    Droller MJ. Markers in bladder cancer. Issues to consider. J Urol. 1998; 160: 2009-2010.
  • 15
    Herr HW. Tumor progression and survival of patients with high grade, noninvasive papillary (TaG3) bladder tumors: 15-year outcome. J Urol. 2000; 163: 60-61.
  • 16
    Eble JN, Epstein JI, Sesterhenn I. World Health Organization Classification of Tumors. Pathology and Genetics of Tumors of the Urinary System and Male Genital Organs. Lyon, France: IARC Press; 2004: 89-158.
  • 17
    Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2(−ΔΔC[T]) methods. Methods. 2001; 25: 402-408.
  • 18
    Ergun S, Kilik N, Ziegeler G, et al. CEA-related cell adhesion molecule 1: a potent angiogenic factor and a major effector of vascular endothelial growth factor. Mol Cell. 2000; 5: 311-320.
  • 19
    Bashar H, Urano T, Fukuta K, et al. Plasminogen activators and plasminogen activator inhibitor 1 in urinary tract cancer. Urol Int. 1994; 52: 4-8.
  • 20
    Pavey SJ, Hawson GA, Marsh NA. Impact of the fibrinolytic enzyme system on prognosis and survival associated with non-small cell lung carcinoma. Blood Coagul Fibrinolysis. 2001; 12: 51-58.
  • 21
    Palmirotta R, Ferroni P, Savonarola A, et al. Prognostic value of pre-surgical plasma PAI-1 (plasminogen activator inhibitor-1) levels in breast cancer. Thromb Res. 2009; 124: 403-408.
  • 22
    Torii K, Kimura H, Li X, et al. Diabetic nephropathy and plasminogen activator inhibitor 1 in urine samples. Rinsho Byori. 2004; 52: 506-512.
  • 23
    McMahon B, Kwaan HC. The plasminogen activator system and cancer. Pathophysiol Haemost Thromb. 2007; 36: 184-194.
  • 24
    Sternlicht MD, Dunning AM, Moore DH, et al. Prognostic value of PAI1 in invasive breast cancer: evidence that tumor-specific factors are more important than genetic variation in regulating PAI1 expression. Cancer Epidemiol Biomarkers Prev. 2006; 15: 2107-2114.
  • 25
    Kuhn W, Schmalfeldt B, Reuning U, et al. Prognostic significance of urokinase (uPA) and its inhibitor PAI-1 for survival in advanced ovarian carcinoma stage FIGO IIIC. Br J Cancer. 1999; 79: 1746-1751.
  • 26
    Chen SC, Henry DO, Hicks DG, Reczek PR, Wong MK. Intravesical administration of plasminogen activator inhibitor type-1 inhibits in vivo bladder tumor invasion and progression. J Urol. 2009; 181: 336-342.
  • 27
    Costantini V, Sidoni A, Deveglia R, et al. Combined overexpression of urokinase, urokinase receptor, and plasminogen activator inhibitor-1 is associated with breast cancer progression: an immunohistochemical comparison of normal, benign, and malignant breast tissues. Cancer. 1996; 77: 1079-1088.
  • 28
    Christiansen JJ, Rajasekaran AK. Reassessing epithelial to mesenchymal transition as a prerequisite for carcinoma invasion and metastasis. Cancer Res. 2006; 6: 8319-8326.
  • 29
    Bajou K, Maillard C, Jost M, et al. Host-derived plasminogen activator inhibitor-1 (PAI-1) concentration is critical for in vivo tumoral angiogenesis and growth. Oncogene. 2004; 23: 6986-6990.
  • 30
    Devy L, Blacher S, Grignet-Debrus C, et al. The pro- or antiangiogenic effect of plasminogen activator inhibitor 1 is dose dependent. FASEB J. 2002; 16: 147-154.
  • 31
    Frandsen TL, Holst-Hansen C, Nielsen BS, et al. Direct evidence of the importance of stromal urokinase plasminogen activator (uPA) in the growth of an experimental human breast cancer using a combined uPA gene-disrupted and immunodeficient xenograft model. Cancer Res. 2001; 61: 532-537.