Although there is no doubt that bisphosphonates (BPs), specific inhibitors of osteoclasts, are beneficial for the treatment of bone metastases, their effects on visceral metastases are unclear. The effect of zoledronic acid (ZOL) was examined in vivo on lung metastasis progression and animal survival, and in vitro on the cellular mechanisms involved.
An animal model of lung metastasis was developed in C3H/He mice inoculated intravenously with a spontaneous murine osteosarcoma POS-1 cell line. Lung metastasis was determined at the time of autopsy. ZOL was assessed in vitro on POS-1 cell proliferation, cell cycle progression, and caspase-1 and -3 activities.
The overall survival in five independent experiments (two series treated with ZOL 0.1 mg/kg twice a week, and three series with 0.1 mg/kg five times a week) showed a significant increase of the actuarial survival: 0.422 ± 0.07 in ZOL-treated animals versus 0.167 ± 0.07 in controls (P = 0.036). Lung metastases were absent in all ZOL-treated mice that survived more than 21 days postinjection as revealed by macroscopic and histologic analysis. In vitro, a 48-hour incubation with 10 μM ZOL inhibited POS-1 cell line proliferation associated with cell cycle arrest in S-phase. In addition, ZOL induced a weak increase of caspase-3 activity, but not caspase-1.
Bisphosphonates (BPs) are currently the most important class of inhibitors of osteoclast-mediated bone resorption. They are widely and successfully used for the treatment of skeletal diseases, such as Paget disease, postmenopausal osteoporosis, and tumor-induced osteolysis.1 BPs have a high affinity for hydroxyapatite mineral in bone and are taken up selectively and adsorbed to mineral surfaces at sites of increased bone turnover, where they inhibit osteoclast activity.2 In addition to their potent antiosteoclast effects, recent preclinical studies have shown that BPs induce apoptosis of cancer cells from several origins, including human myeloma, breast, and prostate carcinoma cell lines.3 It has also been demonstrated that BPs inhibit cancer cell invasion and angiogenesis.3 In spite of the widely recognized beneficial effects of BPs on bone metastases, the effects of BPs on visceral organs are unclear.4 In preclinical studies, the positive effects of zoledronic acid (ZOL) on nonskeletal metastases have been demonstrated in breast carcinoma models.5, 6 Therefore, it is necessary to extend these studies to nonosseous metastases secondary to cancers from other origins. Osteosarcoma (OS) is the most frequent primary bone tumor that develops mainly in the young, the median age of diagnosis being 18 years. A preference for pulmonary metastases compared with other metastatic sites is a distinct feature of OS and 5-year survival rates after the detection of lung metastasis are less than 30%.7 Despite recent improvements in chemotherapy and surgery, the problem of nonresponse to chemotherapy remains and current strategies for the treatment of high-grade osteosarcoma fail to improve its prognosis.8 Therefore, development of new therapies is needed.
In the present study, we investigated the effect of ZOL, an N-BP of the third generation, on the outcome of lung metastases induced by intravenous (i.v.) inoculation of POS-1 osteosarcoma cells. The POS-1 cell line is derived from an osteosarcoma which developed spontaneously in C3H mice. The tumor can be successfully transplanted in C3H mice or POS-1 cells inoculated into the hind footpad of mice and shows spontaneous metastasis to lung.9 Tumors were first recognized macroscopically at 2 weeks after inoculation and developed in more than 90% of inoculated mice at 5 weeks, lung metastasis being observed in all mice that developed tumors. Using the POS-1 cells, we developed a model of pulmonary metastases without primary bone tumor by inoculating the cells i.v. into the tail vein or by a retro-orbital approach. Therefore, this model was used to test the efficacy of ZOL on the progression of pulmonary metastases.
MATERIALS AND METHODS
Four-week-old male C3H/He mice (Elevages Janvier, Le Genest St Isle, France) were housed under pathogen-free conditions at the Experimental Therapy Unit (Medicine Faculty of Nantes, France), in accordance with the institutional guidelines of the French Ethical Committee. The spontaneous murine osteosarcoma cell line POS-1 was kindly provided by the Kanagawa Cancer Centre (Kanagawa, Japan). The cells were cultured in RPMI 1640 medium (BioWhittaker, Verviers, Belgium) supplemented with 10% fetal bovine serum (FBS, Dominique Dutscher, Brumath, France) at 37 °C in a humidified atmosphere (5% CO2/95% air). The mice were anesthetized by inhalation of a mixture of isoflurane/air (1.5%, 1 L/min) combined with an intramuscular injection of Imalgene (100 mg/kg, Merial Laboratories, Lyon, France) prior to i.v. injection of 50 μL of POS-1 cell suspension containing 1.5 × 105 cells. Under these conditions, pulmonary metastases developed rapidly, leading to the death of the animals in 3 weeks after POS-1 cell injection.
Treatment of Mice with Zoledronic Acid
To determine the effect of zoledronic acid (ZOL, kindly provided as the disodium salt by Pharma Novartis, Basel, Switzerland) on lung metastasis development and mouse survival in the POS-1 osteosarcoma model, 24 mice were injected with POS-1 osteosarcoma cells as described above. At Day 2 after tumor cell inoculation, 6 mice were treated with vehicle alone (phosphate buffered saline, PBS), and 18 with ZOL in PBS at 3 different concentrations and sequences: 1) ZOL 100 μg/kg, twice a week; 2) ZOL 100 μg/kg 5 times a week; and 3) ZOL 1 mg/kg, twice a week. Treatment continued until each animal showed signs of morbidity, which included cachexia or respiratory distress, at which point they were sacrificed by cervical dislocation. Lung tumor dissemination was assessed by analyzing the number of tumor foci. Five independent experiments were performed.
Lungs were fixed in 10% buffered formaldehyde, then embedded in paraffin. Sections (5 μM thick) were mounted on glass slides and stained with hematoxylin-eosin-safran (HES) and picrosirius red for collagen.
In Vitro Analyses
Replicate subconfluent cell cultures of POS-1 cells in 96-well plates were treated for 1–3 days with increasing concentrations of ZOL (10−7 to 10−4 M, diluted in RPMI). Cell viability was determined by the sodium 3′[1-(phenylaminocarbonyl)-3,4-tetrazolium]-bis(4-methoxy-6-nitro)benzene sulfonic acid hydrate (XTT) cell proliferation reagent assay kit (Roche Molecular Biomedicals, Mannheim, Germany).
POS-1 cells (2 × 104) grown in 24-well plates were treated with 1 or 10 μM ZOL for the indicated times, washed once with PBS, and lysed with 50 μL of RIPA buffer for 30 minutes. The cells were then scraped off and the protein amount was quantified using the BCA (bicinchominic acid + Copper II sulfate) test (Pierce Chemical, Rockford, IL). Caspase-1 and -3 activity was assessed on 10 μL of cell lysate with the CaspACE assay kit (Promega, Madison, WI) following the manufacturer's recommendations. Cells treated with UV light for 30 seconds 24 hours before harvesting were used as a positive control for caspase activity.
Cell cycle analysis
Confluent POS-1 cells (treated with increasing concentrations of ZOL for 24 and 48 hr) were removed from culture dishes by trypsinization, washed twice in PBS, and incubated in PBS containing 0.12% Triton X-100, 0.12 mM EDTA, and 100 μg/mL DNase-free ribonuclease A (Sigma Chemical, St. Louis, MO). Then, 50 μg/mL propidium iodide (Sigma) were added for each sample for 20 minutes at 4 °C in the dark. The stained nuclei were analyzed by flow cytometry (FACScan, BD Biosciences, Franklin Lake, NJ) using CellQuest software. Cell cycle distribution was based on 2N and 4N DNA content.
Cell proliferation data are expressed as mean ± SE. Comparison between groups was performed by the Mann-Whitney U-test. The effect of ZOL on disease-free survival was determined using the log rank test.
ZOL Increases Mice Survival by Inhibiting Pulmonary Metastases Development
Five independent experiments were performed in which mice received 1.5 × 105 POS-1 cells, leading to the development of lung metastasis in 80–90% of mice, with death occurring between Days 17 and 24 after cell inoculation. First, a dose–response of ZOL was performed to determine the optimal sequence and concentration able to affect animal survival. ZOL was well tolerated, without any overt clinical signs of adverse effects. The results presented in Figure 1A reveal that the three sequences improved survival: ZOL 0.1 and 1 mg/kg twice a week induced a survival rate of 40% 45 days after POS-1 cell inoculation. ZOL 0.1 mg/kg 5 times per week maintained survival at 83% up to day 31, and then 66% at 45 days after POS-1 cells injection. A representation of the overall survival rate of five independent experiments (two series treated with 0.1 mg/kg twice a week, and three series with 0.1 mg/kg five times a week) shows a significant increase of the actuarial survival: 0.422 ± 0.07 in ZOL-treated animals versus 0.167 ± 0.07 in controls (P = 0.036, Fig. 1B).
A high incidence of pulmonary metastases was observed in POS-1 cells-inoculated mice: 87% of mice were positive, with more than 50 tumor foci in each case.
The pulmonary tissue was invaded by tumor foci characterized by high-grade proliferating tumor cells and numerous venous emboli (Fig. 2A,B). Lung metastases were absent in all ZOL-treated mice that survived more than 21 days postinjection, as revealed by macroscopic observation. No metastases were observed at the histologic level 45 days after injection (Fig. 2C,D). The treated tissue is characterized by heterogeneous lung alveolar parenchyma and small fibrotic areas corresponding to regenerative healing tissue (Fig. 2C,D). The presence of collagen in this regenerative tissue was confirmed by picrosirius red staining (data not shown).
ZOL Inhibits POS-1 Cell Proliferation In Vitro and Induces Caspase-3 Activation
To determine whether the in vivo antitumor activity of ZOL could be mediated by a direct effect on POS-1 cell proliferation, ZOL effects were assessed in vitro. The XTT viability test showed that ZOL inhibited POS-1 cell proliferation, with an IC50 value of 44.28 μM (Fig. 3A). Hoechst 33258 staining and caspase activation were investigated to see whether the ZOL-induced inhibition of POS-1 cell proliferation was caused by apoptosis. No modification of nuclear morphology that is characteristic of apoptosis was observed in ZOL-treated POS-1 cells after Hoechst staining (not shown). The results showed that ZOL did not induce any activation of caspase-1 in POS-1 cells (not shown), but it did increase caspase-3 activity at a concentration of 10 μM (72 hr), as compared to positive (UV-treated cells) and negative controls (Fig. 3B).
ZOL Induces S-phase Arrest in POS-1 Cells
Flow cytometry analysis of DNA content was performed with osteosarcoma POS-1 cells to identify cell cycle perturbations after ZOL treatment for 48 hours. The results presented in Figure 4 show a 1.8-fold increase in the number of cells arrested in S-phase after ZOL treatment (the number of cells in S-phase increase from 14% in control cells to 19% and 25% in the presence of 1 and 10 μM ZOL, respectively). This observation was concomitant with a reduction of cells in the G0/G1 phase: 55% and 53% for 1 and 10 μM ZOL, respectively (48 hr incubation) versus 63% in the control untreated POS-1 cells (Fig. 4).
Using a rat model of osteosarcoma, we previously demonstrated that ZOL was able to reduce primary tumor growth and prolong rat survival by decreasing lung metastases dissemination associated with a primary bone tumor.10 However, a direct effect on pulmonary metastases alone could not be identified in the rat model. Using a murine model of lung metastases induced by i.v. injection of osteosarcoma cells, we here demonstrate that ZOL significantly reduced lung metastasis progression, thus extending animal survival. The overall available data on the effects of BPs on visceral metastases in clinical and preclinical studies remain controversial. In a human study, Diel et al.11 initially described that clodronate had adjuvant inhibitory effects on metastases in visceral organs in breast cancer. Later, however, they found no significant effects of clodronate on visceral metastases in the same populations of patients in the extended follow-up.12 Moreover, McCloskey et al.13 did not observe adjuvant effects of clodronate. In contrast, Saarto et al.14 reported that adjuvant treatment with clodronate increased the development of nonskeletal metastases in breast carcinoma patients. Thus, the evidence from the completed clinical trials remain conflicting, as also revealed by preclinical studies in experimental animal models. Indeed, some data suggest that BPs may increase tumor burden and metastases in soft tissues.15, 16 In contrast, the experimental bisphosphonate YH529 reduced nonosseous metastases in the MDA-MB-231 model of breast carcinoma in nude mice.17 More recently, Michigami et al.5 reported that ibandronate reproducibly reduced bone metastases in two animal models of breast carcinoma, but in one of these models (the 4T1 mouse model), neither the preventive nor therapeutic administration of ibandronate caused any effects on lung metastases. In the MDA-MB-231 model of breast cancer, therapeutic administration of ibandronate showed no effects on adrenal metastases.5 More recently, using the 4T1 mouse model, Hiraga et al.6 demonstrated that zoledronic acid significantly suppressed lung and liver metastases and prolonged overall survival of tumor-bearing mice. Using the same model, Nobuyuki et al.18 showed that i.v. ZOL decreased tumor burden not only in bone but also in the liver and lungs of treated mice. In another model of mammary carcinoma cells injected into the medullar space of the proximal tibia of Fisher rats, alendronate reduced lung nodule counts by 95%.19 Here, we confirm the antitumor effect of ZOL on lung metastases progression using an experimental model different from breast carcinoma-derived lung metastases. In the present study, no metastases could be observed macroscopically 21 days after injection, and histologically 45 days postinjection. It can be suggested that micrometastases were present at 21 days, inducing the death of some animals between Days 21 and 45 (Fig. 1), and that at Day 45 the alive animals do not exhibit any further lung metastases. At that time, necrosis could not be observed in ZOL-treated animals. To explain why all of these treated animals do not survive, it can be suggested that in vivo some tumor cells escape to ZOL-induced inhibition of proliferation, probably by developing resistance to ZOL treatment. This phenomenon has been reported in the case of myeloma cells treated with alendronate, another N-BP.20 In this study, although N-BP induced apoptosis of myeloma cells in vitro, most in vivo studies fail to demonstrate a corresponding antitumor effect. This discrepancy might reflect the development of a metabolic resistance to the antitumor effect of N-BP in myeloma cells when they are exposed to N-BP for a prolonged time. In our laboratory, we developed a rat model of osteosarcoma in vivo and used the corresponding cell lines for in vitro experiments.10 When these cells were maintained for a long time in culture, part of these cells became resistant to ZOL treatment, so we can suggest that the same phenomenon happened with POS-1 cells.
Thus, ZOL, which seems to be one of the most effective antiresorptive BPs in vitro, may also exert potent antitumor activity in vivo against the progression of visceral metastases. Inhibition of several cellular mechanisms such as vascularization, adhesion, invasion, and migration have been proposed to explain this phenomenon.3 These mechanisms could be studied with the POS-1 cells used to induce pulmonary metastases in the present model. Indeed, the availability of the corresponding cells will allow testing the effects of ZOL on migration and invasion of POS-1 cells in vitro, and also integrin expression and cell adhesion to endothelial cells.
The ZOL doses used in the present study are justified, as 0.1 mg/kg is clearly equivalent to the clinical dose (4 mg i.v. every 3–4 weeks is equivalent to approximately 100 μg/kg of the research grade disodium salt). However, even if a dosing frequency of twice or five times a week is much greater, it could be justified by the very aggressive nature of lung metastases in the murine model and the short survival times. The doses and schedules of ZOL used in this preclinical study are in agreement with the subchronic and chronic toxicity data given by the Novartis Pharma Laboratories (pers. commun.). Therefore, ZOL doses used in the present study are related to the achievable, safe levels in humans. The observation that lung metastases were reduced suggests a direct effect of ZOL on tumor cells that was not dependent on its effect on the bone microenvironment. However, given the low transient levels of ZOL in blood and soft tissues, we cannot discriminate between a direct effect on tumor cell dissemination from the primary injection site or growth inhibition of the secondary lung metastases. Among the mechanisms hypothesized for the BP antitumor activity, data from the literature report that caspase-dependent apoptosis appears to be the major mechanism responsible for BP-induced tumor cell apoptosis, and caspase-3 is certainly the major player in this response.21 In our experimental model, ZOL caused a direct inhibition of POS-1 cell proliferation and an accumulation of cells in the S-phase of the cycle. Similar results were observed by Evdokiou et al.22 Indeed, using a panel of human osteogenic sarcoma cell lines, those authors demonstrated that ZOL reduced cell numbers in a dose- and time-dependent manner, due either to cell cycle arrest in the S-phase or to the induction of apoptosis.22 A comparable mechanism was also described for N-containing BPs such as ZOL in other cell types—for example, melanoma cells.23 In our study, cell cycle arrest in the S-phase was accompanied by a weak activation of caspase-3 activity, a well-characterized effect of N-BPs.21 However, it is difficult to conclude which is the exact mechanism involved in this model, as Hoescht staining did not reveal any modification of nuclear morphology in ZOL-treated cells (not shown). Moreover, in the cell cycle analysis (Fig. 4), no cells in the sub-G0/G1 phase were observed that could represent apoptotic cells. Therefore, the weak caspase-3 activation observed in our study may be nonspecific, associated with a general cytotoxic effect of the high concentration of ZOL. TUNEL staining, performed on rat osteosarcoma tumor samples, were negative for ZOL-treated animals. Another mechanism could be envisaged, such as anoikis, previously reported in human osteogenic sarcoma cells treated with ZOL by Evdokiou et al.22
In conclusion, using a model of lung metastases different in its origin from the well-studied breast carcinoma, we confirm the direct antitumor effect of ZOL both in vitro (antiproliferative and apoptotic effect) and in vivo on lung metastasis progression, leading to a significant prolongation of disease-free survival in animals. This result reveals that this compound could benefit patients with nonskeletal metastases, including osteosarcoma patients who remain at high risk of eventual relapse, with overt metastatic disease, with tumors that recur after treatment, or that show a low degree of necrosis after administration of chemotherapy and continue to have an unsatisfactory outcome.
The authors thank Dr. Jonathan Green for helpful discussions and C. Bailly, A. Hivonnait, and C. Le Corre from the Experimental Therapy Unit of the IFR26 (Nantes, France) for technical assistance.