Tumor necrosis factor–related apoptosis-inducing ligand (TRAIL) plays an important role in the process of lymphocyte-mediated cytotoxicity against malignant cells. Osteoprotegerin (OPG) is a soluble decoy receptor for TRAIL, and circulating OPG has been implicated in the protection of cells from TRAIL-mediated apoptosis. Thus, OPG may protect tumor cells from lymphocyte-mediated cytotoxicity and, as a result, contribute to tumor progression. In the current study, the authors investigated this hypothesis in patients with bladder carcinoma.
Serum OPG levels for 185 patients with bladder carcinoma were determined using an enzyme-linked immunosorbent assay. These levels then were assessed for potential correlations with various disease characteristics and outcome measures.
The mean serum OPG concentration in patients with bladder carcinoma was approximately 3 times greater than the mean concentration in healthy individuals, and among patients with bladder carcinoma, higher tumor stage and grade were found to be associated with increased serum OPG levels. Within the subpopulation of patients with superficial bladder carcinoma, after a follow-up period of 5 years, those who had low serum OPG levels tended to have a longer postoperative tumor-free interval compared with those who had high serum OPG levels. Furthermore, among patients with muscle-invasive bladder carcinoma, the 5-year disease-specific survival rate was greater for those who had low serum OPG levels than for those who had high serum OPG levels.
Tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL) and its receptors participate in the cytotoxic and apoptotic mechanisms mediated by cytotoxic T lymphocytes (CTL) and natural killer (NK) cells.1, 2 TRAIL is a member of the TNF family and has been shown to induce apoptosis in a variety of malignant cells, including bladder carcinoma cells.3, 4 TRAIL-induced apoptosis appears to be restricted to malignant cells, although the mechanisms underlying this selectivity are unclear.5, 6 TRAIL mediates apoptosis via two membrane-bound receptors, DR4 and DR5, both of which contain cytoplasmic death domains.7, 8 Two additional cell surface receptors, DcR1 and DcR2, also bind TRAIL; however, these receptors do not induce apoptosis and instead function as decoys.9, 10
Recent studies have demonstrated that TRAIL also binds to osteoprotegerin (OPG) and that OPG, a member of the TNF receptor family, acts as another soluble decoy receptor for TRAIL.11, 12 Thus, OPG may compete with other TRAIL receptors on target cell surfaces and thereby inhibit TRAIL-mediated apoptosis. It is therefore conceivable that immune-mediated antitumor cytotoxicity, which relies on (among other mechanisms) the binding of TRAIL to death receptors, may be inhibited by OPG. Such inhibition may allow tumor cells to escape immune surveillance. Consequently, circulating OPG may play a critical role in the process of tumor progression.
Notably, elevated serum OPG levels have been observed in patients with several different types of hematologic and nonhematopoietic malignancies.13, 14 Previous studies have reported the prognostic significance of circulating cytotoxic lymphocytes directed against autologous tumor cells in patients with bladder carcinoma.15 Furthermore, the prognostic significance of soluble Fas and Fas ligand levels in serum samples from patients with bladder carcinoma has been documented.16, 17 Thus, we hypothesized that serum OPG levels, like serum Fas and Fas ligand levels, may also possess prognostic significance for patients with bladder carcinoma. The current study was designed to test this hypothesis.
MATERIALS AND METHODS
Peripheral blood samples were obtained from 185 patients with initial primary bladder carcinoma who had not yet undergone surgery or received any other type of anticancer therapy. These patients included 150 men and 35 women, who ranged in age from 24 years to 89 years. Histologic diagnosis indicated that all patients had transitional cell carcinoma of the bladder. Fifty-three patients had UICC TNM (2002) Grade 1 tumors, 66 had Grade 2 tumors, and 66 had Grade 3 tumors. The TNM status distribution was as follows: Tis, n = 10; Ta, n = 91; T1, n = 45; T2, n = 10; T3, n = 16; and T4, n = 4; N1–3, n = 5; and M1, n = 4. All documented metastases occurred in the lung. Blood samples also were collected from 41 healthy donors who had no history of malignant disease. At collection, all blood samples were confirmed to be negative for findings indicative of confounding diseases or conditions. Informed consent was obtained from all study participants.
Serum was isolated via centrifugation of blood samples, and all isolated serum samples were frozen and stored at −80 °C for future enzyme-linked immunosorbent assays (ELISAs).
ELISA for OPG
A sandwich ELISA performed according to the manufacturer's protocol (Immundiagnostik, Bensheim, Germany) was used to measure serum OPG levels. All OPG concentration measurements were calibrated against titration curves that were generated using reference standards. Using this method, it was possible to ascertain serum OPG levels in excess of 0.14 pM. Repeat measurements yielded consistent results.
Patients were divided into two groups on the basis of serum OPG levels. Patients with ‘high’ levels had serum OPG concentrations that exceeded the median value, and patients with ‘low’ levels had serum OPG concentrations that were less than the median value.
Fresh tumor cells obtained from patients with bladder carcinoma were separated from surgical specimens for in vitro primary culturing, as is described elsewhere.18, 19 In brief, cell suspensions were prepared by treating finely minced tumor tissue samples with collagenase (concentration, 3 mg/mL; Sigma Chemical Co., St. Louis, MO). After being washed with RPMI-1640 medium (GIBCO Biocult, Glasgow, United Kingdom), cell suspensions were layered on discontinuous gradients consisting of 2 mL 100% Ficoll-Hypaque (Pharmacia, Uppsala, Sweden) and 2 mL 80% Ficoll-Hypaque in 15 mL plastic tubes and centrifuged at 400g for 30 minutes. Lymphocyte-rich mononuclear cells were collected from the 100% Ficoll-Hypaque interface, and tumor and mesothelial cells were collected from the 80% Ficoll-Hypaque interface. In some cases, cell suspensions enriched with tumor cells were contaminated with monocyte/macrophages, mesothelial cells, or lymphocytes. To prevent further contamination of host cells, we layered all cell suspensions on discontinuous gradients consisting of 2 mL each of 25%, 15%, and 10% Percoll (Amersham, Little Chalfont, United Kingdom) in complete medium in 15 mL plastic tubes and centrifuged these suspensions at 25g for 7 minutes at room temperature. Tumor cells that had been separated from lymphoid cells were collected from the bottom of the tube, washed, and suspended in complete medium. Morphologic examination of Wright–Giemsa-stained smears revealed that in most cases, contaminating nonmalignant cells accounted for < 5% of these tumor cell samples, which typically were > 93% viable according to the trypan blue dye exclusion test. Samples with < 5% contamination were accepted for use in the current analysis.
Tumor cells were suspended in RPMI-1640 medium supplemented with 25 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, 2 mM L-glutamine, 1% nonessential amino acids, penicillin at a concentration of 100 units per mL, 100 μg/mL streptomycin, and 10% heat-inactivated fetal bovine serum (all from GIBCO Biocult); hereafter, this solution is referred to as complete medium. Tumor cells (concentration, 2 × 105 cells per mL) were incubated in complete medium for 3 days at 37 °C in a humidified atmosphere containing 5% CO2, after which the culture medium was collected and centrifuged. Supernatants then were frozen and stored at −80 °C for future ELISAs.
The human bladder carcinoma cell lines T24, J82, and HT1197 were maintained in complete medium as monolayers on plastic dishes.20, 21
Measurement of Direct Cytotoxicity Using the 3-(4,5-Dimethylthiazol-2-yl)-2,5-Diphenyltetrazolium Bromide (MTT) Assay
The MTT assay was used to assess direct tumor cell lysis, as has been described previously.22, 23 In brief, 100 μL of an HT1197 bladder carcinoma cell suspension (2 × 104 cells) was added to each well of a set of 96-well flat-bottomed microtiter plates (Corning Glassworks, Corning, NY), and each plate was incubated for 24 hours at 37 °C in a humidified atmosphere containing 5% CO2. After incubation, the supernatant was aspirated, tumor cells were washed 3 times with RPMI medium, and 200 μL TRAIL plus serum or culture supernatant was aliquotted into 96-well plates. Each plate was incubated for 3 days at 37 °C. Following this incubation step, 20 μL MTT working solution (concentration, 5 mg/mL; Sigma Chemical Co.) was added to each culture well, and cultures were subsequently incubated for 4 hours at 37 °C in a humidified atmosphere containing 5% CO2. The culture medium then was removed from the microtiter wells and replaced with 100 μL isopropanol (Sigma Chemical Co.) supplemented with 0.05 N HCl. The absorbance of each well at 540 nm was measured using a microculture plate reader (Immunoreader; Japan Intermed Co., Tokyo, Japan). Percent cytotoxicity was calculated using the following formula: % cytotoxicity = [1 − (absorbance of experimental wells/absorbance of control wells)] × 100.
All measurements were made in triplicate. Disease-specific survival rates and postoperative tumor-free intervals were calculated using the Kaplan–Meier method. The generalized Wilcoxon test and the Cox–Mantel test were used to evaluate statistical differences in survival rate and tumor-free interval between patients with high serum OPG levels and patients with low serum OPG levels. The Student t test and the Pearson correlation test were used for all other statistical analyses. P values ≤ 0.05 were considered indicative of statistical significance.
Circulating OPG Levels in Serum from Healthy Individuals and in Serum from Patients with Bladder Carcinoma
ELISA was used to evaluate serum OPG levels in samples obtained from healthy control individuals (n = 41) and from patients with bladder carcinoma (n = 185). The mean serum OPG concentrations in these two groups were 4.1 and 12.9 pM, respectively (Fig. 1). Thus, OPG levels were elevated by a factor of approximately 3 in serum samples obtained from patients with bladder carcinoma. A relatively high serum OPG concentration (> 10 pM) was noted in 1 healthy control individual; however, this individual was negative for all other confounding diseases and conditions.
Serum OPG Levels in Patients with Bladder Carcinoma According to Histologic Stage and Grade
Serum OPG levels in patients with bladder carcinoma were analyzed according to histologic stage and tumor grade. Patients with muscle-invasive (T2–4N0M0) bladder carcinoma were found to have significantly higher serum OPG concentrations compared with patients who had superficial (Tis, Ta, or T1N0M0) disease (Fig. 2). Furthermore, serum OPG levels were significantly elevated in patients with T1 bladder carcinoma relative to patients with Ta disease and in patients with metastatic bladder carcinoma relative to patients with muscle-invasive disease. With regard to tumor grade, serum OPG levels were significantly higher in patients with Grade 2 bladder carcinoma compared with patients who had Grade 1 disease (Fig. 3). Likewise, serum OPG levels in patients with Grade 3 bladder carcinoma were significantly higher than those observed in patients with Grade 2 disease. These results indicate that serum OPG levels tend to increase with increasing disease stage and increasing tumor grade, a finding that is supported by preliminary experiments demonstrating that postsurgical serum OPG levels are significantly lower than serum OPG levels measured before curative surgery (data not shown).
Correlation between Serum OPG Levels and Postoperative Tumor-Free Period in Patients with Superficial (Ta or T1) Bladder Carcinoma
The postoperative clinical courses of patients with superficial (Ta or T1) bladder carcinoma who underwent transurethral resection were retrospectively evaluated. Postoperative tumor-free intervals were estimated using the Kaplan–Meier method. For the purposes of this subanalysis, patients with superficial bladder carcinoma were divided into two groups—those with high serum OPG levels (i.e., OPG levels greater than the median value) and those with low serum OPG levels (i.e., OPG levels less than the median value). Within the subpopulation of patients with superficial bladder carcinoma, after a follow-up period of 5 years, those who had low serum OPG levels tended to have a longer postoperative tumor-free interval compared with those who had high serum OPG levels (Fig. 4). This finding suggests that serum OPG concentration may be a significant prognostic parameter for patients with Ta or T1 bladder carcinoma.
Correlation between Serum OPG Level and Postoperative Clinical Course in Patients with Muscle-Invasive (T2–4) Bladder Carcinoma
The postoperative clinical courses of patients with muscle-invasive (T2–4N0M0) bladder carcinoma who underwent radical cystectomy also were retrospectively evaluated using the Kaplan–Meier method. In this subanalysis, again, patients were divided into two groups—those with high serum OPG levels (i.e., OPG levels greater than the median value) and those with low serum OPG levels (i.e., OPG levels less than the median value). Among patients with muscle-invasive bladder carcinoma, the 5-year disease-specific survival rate was found to be higher for those with low serum OPG levels than for those with high serum OPG levels (Fig. 5). This finding suggests that OPG concentration may also be a significant prognostic factor for patients with muscle-invasive bladder carcinoma, with low serum OPG levels being considered a favorable prognostic indicator.
Effects of Sera Obtained from Patients with Bladder Carcinoma and of Bladder Carcinoma Cell Culture Supernatants on TRAIL-Mediated Cytotoxicity against HT1197 Cells
It has been reported elsewhere that OPG inhibits TRAIL-mediated cytotoxicity.11, 12 We attempted to investigate this inhibitory effect by assessing the extent to which serum from patients with bladder carcinoma and bladder carcinoma cell culture supernatants were able to block TRAIL-mediated cell death. Because the HT1197 bladder carcinoma cell line does not secrete OPG (data not shown), TRAIL-mediated cytotoxicity against HT1197 cells was considered strictly indicative of the inhibitory activity of OPG in serum samples and culture supernatants. Serum samples from patients with bladder carcinoma and bladder carcinoma cell culture supernatants both were found to significantly inhibit TRAIL-mediated cytotoxicity (Table 1). These results suggest that OPG that is present in serum and in culture supernatants (as assessed using ELISA) is biologically active and can reduce TRAIL-mediated cytotoxicity.
Table 1. Inhibition of TRAIL-Mediated Cytotoxicity by Serum Samples from Patients with Bladder Carcinoma and Bladder Carcinoma Cell Culture Supernatants
Serum sample/supernatant source
OPG concentration (pM)
Mean % cytotoxicity against HT1197 bladder carcinoma cells ± SDa
In the current study, it was found that serum OPG levels were higher in patients with bladder carcinoma than in healthy volunteers and that serum OPG concentration was positively correlated with disease progression and tumor grade in patients with bladder carcinoma. To our knowledge, the current study was the first to demonstrate that after a 5-year follow-up period, among patients with superficial (Ta or T1) bladder carcinoma, those with low serum OPG levels tended to have longer tumor-free intervals compared with those who had high serum OPG levels. We also found that among patients with muscle-invasive bladder carcinoma, those with low serum OPG levels had a higher 5-year disease-specific survival rate than did those with high serum OPG levels. Although we are reporting on a limited number of patients monitored over a relatively short follow-up period, our preliminary data indicate that serum OPG levels may be a significant prognostic factor for patients with bladder carcinoma.
Several studies have measured OPG levels in serum. With regard to hematologic disorders, serum OPG levels were found to be higher in patients with Hodgkin disease and in patients with non-Hodgkin lymphoma than in healthy volunteers,13 although that same study found no significant difference in serum OPG levels between patients with leukemia and healthy control individuals. Reduced serum OPG levels have been documented in patients with multiple myeloma.13, 24 With regard to solid tumors, serum OPG levels were found to be elevated in patients with colorectal carcinoma, patients with pancreatic carcinoma, and patients with prostate carcinoma compared with healthy individuals.13, 14 In contrast, serum OPG levels were significantly lower in patients with sarcoma compared with healthy donors.13 Also, among patients with malignant disease, a trend toward increased serum OPG levels in patients with metastatic disease compared with patients who had localized disease was documented.13 Likewise, the current study has demonstrated that patients with bladder carcinoma, and especially those with metastatic disease, have elevated serum OPG levels compared with healthy control individuals. Taken together, these findings suggest that serum OPG concentration may be a measure of tumor burden in patients with nonhematopoietic malignancies.
Cell-mediated immunity plays an important role in the process of immune surveillance of bladder carcinoma cells.15, 25 An important parameter of cell-mediated cytotoxicity, the activity of peripheral blood lymphocytes against autologous tumor cells is a significant and independent prognostic indicator for patients with bladder carcinoma.15 Cytotoxic activity against autologous tumor cells is mediated by CTLs and NK cells. Furthermore, TRAIL-induced apoptosis is involved in both CTL-mediated and NK-mediated antitumor cytotoxic mechanisms,1, 2 and OPG has been implicated in the inhibition of TRAIL-mediated apoptosis.11, 12 Thus, elevated levels of OPG in the circulation may reduce cytotoxic activity against autologous tumor cells by inhibiting TRAIL-mediated apoptosis. The current study tested this hypothesis and revealed that serum samples from patients with bladder carcinoma and bladder carcinoma cell culture supernatants, all of which were high in OPG content, inhibited TRAIL-mediated cytotoxicity. This finding indicates that elevated serum OPG levels may be associated with poor prognosis as a result of OPG's inhibitory effect on TRAIL-mediated cytotoxicity. It therefore is reasonable to conclude that elevated serum OPG levels are involved in a novel mechanism by which tumor cells can escape from immune surveillance, with this mechanism promoting tumor progression in patients with bladder carcinoma.
Several features that are potentially associated with mechanisms of resistance to TRAIL-mediated apoptosis have been identified, including reduced expression of TRAIL receptors DR4 and DR5 and enhanced expression of the antagonistic TRAIL receptors DcR1 and DcR2.7–10 The existence of multiple TRAIL receptors suggests an unexpected level of complexity in the regulation of TRAIL-mediated signaling. Antiapoptotic molecules such as Bcl-2 and Bcl-xL may be specifically associated with resistance to TRAIL-mediated apoptosis (among other forms of apoptosis).26, 27 Because TRAIL induces apoptosis in malignant cells in a caspase-dependent manner, resistance to TRAIL-mediated apoptosis may also be dependent on caspase expression levels.28, 29 FLICE-like inhibitory protein has been shown to bind to caspase-8 and prevent the activation of downstream events leading to apoptosis, including TRAIL-mediated apoptosis.30, 31 The current study suggests that OPG production by bladder carcinoma cells may represent another mechanism of resistance to TRAIL-mediated apoptosis. Nonetheless, further studies are required to elucidate the mechanisms by which bladder carcinoma cells acquire resistance to TRAIL-mediated cytotoxicity.
The precise cellular source of OPG has not been elucidated. We speculate that OPG may be derived from malignant cells and/or normal tissue. Previous reports have demonstrated that OPG is produced by prostate carcinoma cells and Hodgkin lymphoma cells.32, 33 Normal tissues, including lung, heart, liver, stomach, intestinal, kidney, skin, brain, spinal cord, thyroid gland, and bone tissue, also have been found to produce OPG.34, 35 In the current study, preliminary experiments revealed that OPG was present in bladder carcinoma cell culture supernatants and also in the supernatants of primary cultures derived from surgical specimens. These findings suggest that both malignant cells and normal tissue produce OPG, although further studies are required to conclusively identify the cellular source of OPG in patients with bladder carcinoma.
Osteoclast differentiation recently was shown to be positively and negatively regulated by a complex signaling system involving receptor activator of nuclear factor κB (RANK), RANK ligand (RANKL), and OPG. Among the key interactions that these molecules take part in are the binding of RANK on osteoclast progenitor cell surfaces to RANKL on osteoblasts during direct cell contact, an interaction that induces osteoclastogenesis and activates bone resorption, and the binding of soluble OPG to RANKL, an interaction that suppresses osteoclastogenesis by interfering with the interaction between RANK and RANKL.36 Such suppression of osteoclastogenesis has been observed at OPG concentrations of 10–100 ng/mL.34, 37 Although some patients with bladder carcinoma have significantly elevated levels of circulating OPG, these levels are not sufficient to suppress osteoclast formation.
The current study demonstrated that serum OPG concentration may have clinical usefulness as a prognostic marker for patients with bladder carcinoma; however, only a limited number of patients with T4, N1, or M1 disease were included in the current cohort, and as a result, broad generalization of the prognostic value of serum OPG levels was not possible. Thus, our findings warrant further investigation and require validation in larger patient populations.
Although overall rates of response to chemotherapy for patients with bladder carcinoma have improved, metastasis and disease recurrence remain major problems. Therefore, novel therapeutic approaches are required for patients with metastatic or recurrent disease. The up-regulation of serum OPG levels in patients with bladder carcinoma (and especially in patients with advanced-stage or high-grade disease) compared with healthy donors suggests that OPG may be a useful molecular target for anticancer therapy. Accordingly, inhibition of the production or biologic activity of OPG may prevent the progression of bladder carcinoma. Furthermore, strategies aimed at inhibiting OPG production may increase the susceptibility of malignant cells to TRAIL-mediated killing by cytotoxic lymphocytes and thus improve clinical outcomes.
In conclusion, the data presented in the current report demonstrate that serum OPG concentration is positively correlated with histologic disease stage and tumor grade in patients with bladder carcinoma and that elevated serum OPG levels are associated with early recurrence in such patients. The observed correlation with postoperative prognosis suggests that serum OPG levels may possess utility as a prognostic marker for patients with bladder carcinoma, and in turn, accurate assessment of prognosis may aid in the selection of patients to receive intensive surgical or chemotherapeutic treatment.
The authors thank Yukako Morioka and Kate Dinh for their assistance in the preparation of the current article.