Evaluation of effects of radiation therapy combined with either pamidronate or zoledronate on canine osteosarcoma cells

Abstract

Canine osteosarcoma is a devastating disease with an overall poor prognosis. Radiation therapy and bisphosphonates are currently used in combination for palliative treatment, despite a paucity of literature that investigates their combined use. The objectives of this study were to assess the in vitro effects of radiation therapy and bisphosphonates on canine osteosarcoma cells when used in combination. Canine osteosarcoma cell lines D17 and Dharma were treated with radiation and pamidronate or zoledronate, both alone and in combination. The effects of these treatments were assessed using clonogenic survival and cell viability assays. Dose-dependent decreases in clonogenic survival and cell viability were observed for both radiation and bisphosphonate treatment. Combination index analysis revealed antagonistic interactions when radiation and bisphosphonates were used in combination at specific doses for both D17 and Dharma osteosarcoma cells. Further investigation of the combined effects of radiation and bisphosphonates for the palliative treatment of canine osteosarcoma is warranted.

Introduction

Radiation therapy (RT) and bisphosphonates (BPs) are common palliative treatment approaches for osteosarcoma (OSA) in dogs that are not undergoing standard-of-care treatment (1–5). The goal of using palliative RT in patients with OSA is to provide pain relief (6–8). External beam radiation is the most common form of RT used and coarse-fractionated protocols that deliver higher doses over a short time period are typically used in a palliative treatment setting (9–13). Radiation therapy relieves pain in dogs with OSA in several ways. It reduces the number of cells (including inflammatory cells) within the bone marrow, which decreases pressure within the affected bone, inhibits osteoclastic bone destruction, and may reduce tumor burden by directly killing OSA cells (8,11,14,15).

In addition to RT, BPs have gained popularity in veterinary medicine over the past decade for the palliative treatment of OSA (1,16). These drugs are often used in combination with RT with the expectation that combining these modalities may result in more palliative benefit than either treatment alone. Bisphosphonates (BPs) are synthetic analogues of inorganic pyrophosphates that inhibit osteoclast function to reduce bone resorption and have direct anti-tumor effects, including inhibiting tumor cell proliferation, adhesion, and invasion, as well as promoting apoptosis (17–25). These drugs have been used for decades to treat osteoporosis, as well as primary and metastatic bone tumors in humans. The goal of BP use in OSA is to provide analgesia through bone antiresorptive properties, to prevent development of bone metastases, and to cause direct anticancer effects (17–21).

Pamidronate (PAM), which is widely used in veterinary medicine, is a second-generation BP that, in addition to its conventional mechanisms of action, has also demonstrated anti-angiogenic effects by reducing levels of vascular endothelial growth factor (VEGF) in serum of human cancer patients (26). In canine OSA, PAM has been shown to inhibit OSA growth through a non-apoptotic mechanism in vitro and to reduce pain and pathological bone turnover in a clinical trial (2,27).

Zoledronate (ZOL), a third generation BP, is the most potent intravenous BP approved for human use (28). In veterinary medicine, use of ZOL to manage canine OSA may be increasing, although there is little evidence for its benefits over other BPs in this context. In vitro studies have shown anti-tumor effects of ZOL in both human and canine OSA cell lines, including decreased cell growth, a dose-dependent increase in apoptosis, alteration of cell cycle distribution, inhibition of tumor cell invasion, and anti-angiogenic effects (29–33). Zoledronate has also been shown to reduce OSA-induced bone lysis in a nude mouse canine OSA orthotopic xenograft model (34). While ZOL may have improved bone pain in this study, however, it was also associated with an increased risk for pulmonary metastases in these mice.

Independently, RT and BP each have the potential to improve cancer-related bone pain. In veterinary oncology, there remains a paucity of studies specifically investigating the combined use of RT and BP, despite the frequent clinical application of this combination treatment approach. Three studies in canine OSA patients have evaluated the combined use of PAM with RT and/or chemotherapy (1,16,35). In each of these studies, PAM failed to improve pain relief or median survival time (MST) when combined with RT and/or chemotherapy. One study noted, however, that chemotherapy either with or without PAM had a similar MST, whereas combined RT and PAM resulted in a decreased MST (1). However, this was a retrospective study with relatively low case numbers in each group and not a prospectively designed controlled clinical trial. Taken together, these studies suggest that there could be a lack of benefit and perhaps a potential negative impact when PAM is combined with RT, but further investigation is needed.

There are currently no reports evaluating the in vitro effects of the combination of PAM and RT on OSA cells. Ryu et al (36) demonstrated, however, that ZOL and RT resulted in significantly more growth inhibition of murine and human OSA cell lines in vitro when used in combination than when used alone. Zoledronate (ZOL) appears to be a radiosensitizer in murine and human OSA cells, which may result in a different effect when ZOL is combined with RT than when PAM is combined with RT (36,37).

Due to the limited in vitro evaluation of the combination of either PAM or ZOL with RT in canine OSA cell lines, the objectives of this study were to evaluate the in vitro proliferation and viability of canine OSA cells when subjected to the combination of either PAM or ZOL with RT. The null hypothesis of this study was that combining either PAM or ZOL with RT to treat canine OSA cells would not significantly increase inhibition of proliferation and viability compared to PAM, ZOL, and RT treatment used alone.

Materials and methods

Cell culture

D17 and Dharma canine OSA cell lines were used for all experiments. D17 cells (38) were derived from a lung metastatic lesion and obtained from Sigma-Aldrich/European Collection of Cell Cultures (ECACC) and Dharma cells were isolated from a primary appendicular lesion in a clinical case and adapted to culture by Dr. Anthony Mutsaers (39). Cells were grown in standard cell culture dishes with Dulbecco’s modified Eagles medium (Hyclone DMEM; Fisher Scientific, Ottawa, Ontario) with 10% fetal bovine serum (FBS; Life Technologies, Burlington, Ontario) and 1% penicillin-streptomycin (PS; BioWhittaker, Mississauga, Ontario). Cells were incubated at 37°C in a 5% carbon dioxide (CO₂) humidified incubator.

Clonogenic survival assay

For each cell line and treatment condition, 6, 6-well cell culture plates were plated at 500 cells/well in 3 mL of media for the D17 cell line and at 2000 cells/well in 3 mL of media for the Dharma cell line. Plates were incubated overnight and cells were then treated with PAM (pamidronate disodium salt hydrate; Sigma-Aldrich, Oakville, Ontario) or ZOL (zoledronic acid; Sigma-Aldrich). Two dose levels of PAM (10 to 30 μM) or ZOL (0.4 to 2 μM) and 1 vehicle control were used, with 2 wells/plate for each dose. Given the difference in potency between the 2 BP drugs, doses were chosen after initial dose-optimization studies. All doses used fall within previously reported ranges used in in vitro studies (27,29,36,37). Media was removed from each well and replaced by 3 mL of BP-containing media or standard culture media for the control wells. Plates were subsequently placed in the incubator before receiving RT later that same day. One plate from each experiment received a single dose of RT (2 to 10 Gy) using a 6-MV linear accelerator at a rate of 400 monitor units per min (Clinac IX System; Varian Medical Systems, Palo Alto, California, USA). Plates were placed between 2 solid water-equivalent plates during radiation. A medical physicist verified this dose distribution. Control plates (0 Gy) remained outside the radiation vault during cell treatment.

Plates were returned to the incubator and colony formation was monitored daily. The experiment was terminated before the control colonies became confluent (7 d) and cells were then fixed and stained with 0.5% crystal violet in 20% methanol. Colonies were counted using light microscopy. A colony was defined as an aggregate of ≥ 50 cells. Overall, each treatment had duplicate wells per experiment and each experiment was conducted 3 times.

Viability assay

For each cell line and treatment condition, 6, 96-well cell culture plates were plated at 500 cells/well in 150 μL of media for the D17 cell line and at 2000 cells/well in 150 μL of media for the Dharma cell line. Plates were placed in the incubator overnight. Cells were subsequently treated in quadruplicate with BP the following morning. Doses were chosen after initial optimization studies. All doses fell within previously reported ranges used in in vitro studies (27,29,36,37). Five doses of either PAM (1 to 100 μM) or ZOL (0.2 to 100 μM) and 1 control were used. Plates were subsequently placed in the incubator before receiving RT later that day. One plate from each experiment received a single dose of RT (2 to 10 Gy) using a 6-MV linear accelerator at a rate of 400 monitor units per min (Clinac IX System; Varian Medical Systems). Plates were placed between 2 solid water-equivalent plates during radiation. A medical physicist verified this dose distribution. Control plates (0 Gy) remained outside the radiation vault during cell treatment.

Plates were returned to the incubator for 7 d. On day 7 after RT, cell viability was assessed using the Resazurin Cell Viability Kit (Sigma Aldrich). Resazurin solution (20 μL) was added to each well and absorbance readings were obtained 6 h later using a Synergy 2 spectrophotometer (BioTek, Winooski, Vermont, USA) at an excitation wavelength of 570 nm and emission wavelength of 600 nm. Overall, each treatment had quadruplicate wells per experiment and each experiment was conducted 3 times.

Statistical analysis

A general linear mixed model [2-way analysis of variance (ANOVA)] was used to test the fixed effects of PAM or ZOL and RT and their interactions. The random effect of the plate was accounted for in the model. A Shapiro-Wilk test and examination of the residuals assessed the data for normality. If the data were not normally distributed, a log transform was applied. A value of P < 0.05 was considered significant.

After calculating the half maximal inhibitory concentration (IC₅₀) for the viability assays using GraphPad Prism (GraphPad Software, La Jolla, California, USA), analysis of the effects of combining PAM or ZOL and RT were assessed using the combination index (CI) method of Chou and Talalay (40). This was done using the program CompuSyn (ComboSyn, Paramus, New Jersey, USA) to quantify the CI. Only doses resulting in < 50% cell growth inhibition in single agent studies were used in this assessment to allow for potentially equal contributions from PAM or ZOL and RT. The combined effects can be additive (effects of individual treatments added together, CI = 1.0), synergistic (effects of combination greater than individual effects added together, CI < 1.0), or antagonistic (effects of combination less than individual effects added together, CI > 1.0) (40).

Results

Pamidronate

Treatment of D17 cells resulted in a dose-dependent reduction in clonogenic survival (Figure 1A) and cell viability (Figure 1B) for both PAM and RT. Statistically significant differences were only identified during the clonogenic survival experiments. When comparing single agent RT at 2 Gy, 4 Gy, and 6 Gy to the combination of 20 μM PAM and the same RT doses, significantly fewer colonies were formed when PAM and RT were combined (Figure 1C). One other significant difference was identified when comparing cells treated with 10 μM PAM alone to cells treated with 10 Gy RT + 10 μM PAM (Figure 1D).

Figure 1.

Clonogenic survival and cell viability for D17 cells treated with pamidronate (PAM) + radiation therapy (RT). A — Dose-dependent decrease in clonogenic survival (n = 2, mean +/− SEM). B — Dose-dependent decrease in cell viability (n = 4, mean +/− SEM). C — Clonogenic survival of D17 cells treated with 20 μM PAM + RT (n = 6, mean +/− SEM). D — Clonogenic survival of D17 cells treated with 10 μM PAM + RT (n = 6, mean +/− SEM). * P < 0.05.

No significant findings were identified for any colonization experiments when Dharma OSA cells were treated with PAM and RT (Figure 2A). Treatment of Dharma cells resulted in a dose-dependent reduction in viability for RT, but not for PAM (Figure 2B).

Figure 2.

Clonogenic survival and cell viability for Dharma cells. A — Dose-dependent decrease in clonogenic survival, pamidronate (PAM) + radiation therapy (RT) (n = 2, mean +/− SEM). B — Dose-dependent decrease in cell viability, PAM + RT (n = 4, mean +/− SEM). C — Dose-dependent decrease in clonogenic survival, zoledronate (ZOL) + RT (n = 2, mean +/− SEM). D — Dose-dependent decrease in cell viability, ZOL + RT (n = 8, mean +/− SEM).

When the combination index was calculated using the viability results of drug combinations, antagonism was identified for all dose combinations of PAM and RT assessed for both the D17 and Dharma cell lines (Table I).

Table I.

Combination index (CI) for D17 and Dharma cells treated with pamidronate (PAM) or zoledronate (ZOL) + radiation therapy (RT).

Cell line	BP	BP dose	RT dose	CI
D17	PAM	5	2	1.71992667
D17	PAM	10	4	1.408856667
Dharma	PAM	4	2	1.239245
Dharma	PAM	8	4	1.055055
D17	ZOL	0.5	2	1.42468
D17	ZOL	1	4	1.75257
Dharma	ZOL	0.4	2	8.758275
Dharma	ZOL	0.8	4	86.171375

^*Data in this table represent dose combinations that resulted in an IC₅₀ or lower for viability assays.

BP — bisphosphonates.

Zoledronate

Treatment of D17 cells resulted in a dose-dependent reduction in clonogenic survival (Figure 3A) and viability (Figure 4A) for both ZOL and RT. Significant results were identified in both assays.

Figure 3.

Clonogenic survival for D17 cells treated with zoledronate (ZOL) + radiation therapy (RT). A — Dose-dependent decrease in clonogenic survival (n = 2, mean +/− SEM). B — A significant decrease in clonogenic survival of D17 cells was observed after treatment with 0.5 μM ZOL + RT (n = 6, mean +/− SEM). C — 2 μM ZOL + RT (n = 4, mean +/− SEM). * P < 0.05.

Figure 4.

Viability for D17 cells treated with zoledronate (ZOL) + radiation therapy (RT). A — Dose-dependent decrease in cell viability (n = 4, mean +/− SEM). B — Significant decrease in viability of D17 cells after treatment with 5 μM ZOL + RT (n = 6, mean +/− SEM). C — 7.5 μM ZOL + RT (n = 8 to 12, mean +/− SEM). D — 10 μM ZOL + RT (n = 12, mean +/− SEM). E — 20 μM ZOL + RT (n = 8 to 12, mean +/− SEM). F — 0.5 μM ZOL + RT (n = 8, mean +/− SEM). G — 1 μM ZOL + RT (n = 16, mean +/− SEM). H — 2.5 μM ZOL + RT (n = 8 to 16, mean +/− SEM). I — 5 μM ZOL + RT (n = 16, mean +/− SEM). * P < 0.05.

When comparing single agent ZOL at 0.5 μM or 2 μM concentrations to the 6 Gy, 8 Gy, or 10 Gy RT doses treated with the same ZOL concentrations, significantly fewer colonies were formed when ZOL and RT were used in combination (Figures 3B and 3C). When comparing single agent RT irradiated at 2 Gy, 4 Gy, 6 Gy, and 8 Gy to the 5 μM, 7.5 μM, 10 μM, and 20 μM ZOL concentration irradiated at the same doses, significantly less cell viability was identified when ZOL and RT were used in combination, with the exception of the combination of 5 μM ZOL and 8 Gy RT (Figures 4B to 4E). When comparing single agent ZOL at 0.5 μM, 1 μM, 2.5 μM, or 5 μM concentrations to the 6 Gy, 8 Gy, or 10 Gy RT doses treated with the same ZOL concentrations, significantly less cell viability was identified when ZOL and RT were used in combination, with the exception of the following combinations: 0.5 μM + 6 Gy, 0.5 μM + 8 Gy, and 5 μM + 6 Gy (Figures 4F to 4I).

No significant findings were identified for any colonization experiments when Dharma OSA cells were treated with ZOL and RT (Figure 2C). Treatment of Dharma cells resulted in a dose-dependent reduction in cell viability for both ZOL and RT (Figure 2D).

When the combination index was calculated using the viability results of drug combinations, antagonism was identified for all dose combinations of ZOL and RT assessed using the cell viability results in both D17 and Dharma cells, with a stronger antagonistic result identified in the Dharma cells (Table I).

Discussion

Despite increasing use of the combination of RT and BP for palliative treatment of canine OSA, there is limited information about the effects of their combined use. Since each modality alone has demonstrated analgesic effects for dogs with OSA, a positive interaction has been assumed when BP and RT are combined. By evaluating the use of PAM or ZOL in combination with RT on canine OSA cells in vitro, we have shown that their interactions may be more complex than anticipated and that synergistic or even additive effects cannot necessarily be assumed.

Similar to what has previously been reported, we found that use of PAM, ZOL, or RT treatment alone resulted in dose-dependent decreases in OSA cell viability and clonogenic survival (27,29). Significant differences were only identified in the D17 cell line, however, which highlights the potential for heterogeneity of OSA treatment response. Despite their similar classification, all OSA cell lines and tumors do not have an identical response to the same treatment protocol. A contributory factor may be the differing growth rates of D17 and Dharma cells in vitro, with Dharma cells having a longer doubling time than D17. In addition, D17 cells were derived from a metastatic lung lesion, while Dharma cells were derived from a primary bone tumor. This fact may be worthy of consideration, as the clinical use of BP and RT in canine OSA is directed at the primary bone tumor rather than at lung metastases. Although significant differences in cell growth inhibition were seen when different BP doses or different RT doses were compared, few combinations of BP and RT resulted in significantly more inhibition of cell growth than any individual treatment.

It is noteworthy that the significant combinations identified in the D17 cell line only resulted in significantly fewer colonies or significantly less viability when compared to 1 of the treatments in the combination, but not the other treatment. In other words, from the significant ZOL combinations, it is clear that 1 treatment of either ZOL or RT played a greater role in a particular combination. For the significant combinations of ZOL and RT in the clonogenic survival assays, RT played a greater role in decreasing colony outgrowth, whereas ZOL made a greater contribution to reducing viability, except at high RT doses (Figures 3 and 4). These results suggest differential contributions of these 2 modalities to the overall anticancer and/or analgesic effects in the palliative treatment of OSA. It is important to recognize, however, that palliation may be achieved by other means than direct cytotoxic effects of OSA cells by BP and RT, which was the only outcome measured in this study. Therefore, combining BP and RT could lead to more effective palliation in patients, based on non-cancer cell cytotoxic mechanisms.

It can be difficult to discern from the data alone the relative contribution of each treatment effect when delivered in combination, aside from specific combinations where 1 treatment modality clearly resulted in increased inhibition of cell growth on its own compared to the other modality. Relative contributions are better measured through combination index (CI) analysis. The combination index is a mathematical model that analyzes the effects of multiple drugs or treatments in order to determine if their relationship is additive (effects of individual treatments added together, CI = 1.0), synergistic (effects of combination greater than individual effects added together, CI < 1.0), or antagonistic (effects of combination less than individual effects added together, CI > 1.0) (40).

The identification of antagonistic interactions when PAM and RT are combined in vitro under specific experimental conditions supports the results of previous clinical studies that suggest either a negative impact or lack of improvement when these treatment modalities are combined (1,16,35). Similar to PAM, antagonistic interactions were identified by CI analysis with ZOL and RT under specific experimental conditions. These results are in contrast to those reported by Ryu et al (36), who identified ZOL as a radiosensitizing agent producing decreased cell viability when combined with RT. The use of different OSA cell lines and/or experimental conditions may have contributed to these contradictory results (36,37). As with any in vitro study, however, other intrinsic biologic effects present in vivo, such as the tumor microenvironment, are not accounted for, which may contribute significantly to the overall response of tumors to BP and RT.

Based on these results, while a dose-dependent anti-growth response to the individual treatments was confirmed, the effects of combining BP and RT remain variable among canine OSA cell lines tested in vitro. Additionally, results suggest the possibility of an antagonistic relationship between BP and RT. Further study is required to investigate potential contributory factors to these results, such as whether the timing of administration of each treatment contributes to the potential for a negative interaction. In our study, cells were exposed to BP for 2 to 4 h before radiation, which may have produced results that differ from BP treatment after RT. In most clinical patients, BPs are given at the same visit as RT, which is why this treatment sequence was chosen. Finally, clinical recommendations regarding the combined use of bisphosphonates (BPs) and radiation therapy (RT) for the palliative treatment of canine OSA cannot be made without further clinical investigation, which would ideally require a randomized controlled trial.