Cadmium in bone cement induces necrosis and decreases the viability of residual osteosarcoma cells: A xenograft study

Abstract

Objective

The aim of this study was to show whether local application of cadmium-impregnated bone cement can induce apoptosis and decrease the viability of residual osteosarcoma (OS) cells in nude mice.

Methods

K7M2 tumorigenic OS cell line was cultivated in vitro. The xenograft tumor model was formed by subcutaneously adding the tumor cells to athymic nude mice. Tumor was formed within 1 month. Then, mice were randomly assigned to five groups, each containing seven nude mice: control (group 1), wide resection (group 2), intralesional resection (group 3), intralesional resection + bone cement (group 4), and intralesional resection + cadmium embedded in bone cement (group 5). Tumor resection with 1 cm surgical margins was performed in the wide resection group. In intralesional resection groups, tumor tissue was resected with positive margins aiming to leave 15 mm3 of macroscopic tumor tissue. In group 3, the defect was left empty; groups 4 and 5 received bone cements prepared with saline and cadmium solutions, respectively. After the resection, mice were observed for 15 days and sacrificed. Next, surgical resection sites were evaluated histopathologically in each group.

Results

Recurrent tumor was formed in all mice in the wide resection group, and apparent progression of residual tumor was observed in groups 3 and 4. On the contrary, only a thin layer of residual tumor was observed around the bone cement in group 5. Histological evaluation revealed remarkable necrosis in group 5 and lowest viability compared to other groups. No systemic toxic effect related to cadmium was observed.

Conclusion

Our data suggest that local application of cadmium in bone cement has a significant potential to increase tumor necrosis and decrease the viability of residual OS cells.

Osteosarcoma (OS), which is characterized by osteoid production from atypical, malignant osteoblasts, is the most common primary sarcoma of the bone (1). The overall 5-year survival rate for OS is 68%, without significant sex-related difference (2). Complete surgical excision is important to ensure an optimum outcome (3). Tumor staging, the presence of metastases, chemotherapy regimen, anatomic location, size of the tumor, and percentage of tumor cells destroyed after neoadjuvant chemotherapy as well as local recurrence have prognostic effects on the outcome (4). Local recurrence following OS treated with systemic chemotherapy and adequate surgical control of the primary tumor with either a limb-salvage or ablative surgery is one of the toughest and challenging problems for global oncology societies. Once local recurrence occurs, the prognosis gets very poor. Even for such a worst scenario, every attempt should be made to increase the post recurrence survival (PRS). Several factors are important for PRS such as the use of neoadjuvant chemotherapy, obtaining adequate surgical margins in resection of the locally recurrent tumor (preferably amputation, or at least a resection with a 1 cm of clean margin), control of distant metastasis, and longer time interval (preferably > 2 years) between diagnosis and local recurrence (5).

To overcome local recurrence of OS following limb-sparing surgery, local application of different substances, besides chemotherapeutic drugs, especially in bone cement, has received much attention in recent years (6–9). A naturally occurring toxic heavy metal cadmium (Cd) is classified as a human carcinogen (10,11). Cd is also known as an osteotoxin and Cd toxicity may lead to disturbances in calcium metabolism, osteomalacia, osteoporosis, and pathologic fractures (12). Cd is known to induce apoptosis both in vivo and in vitro in numerous cells, including osteoblasts. A remarkable rate of apoptosis was observed in MG-63 OS cells after short-term exposure to Cd (13).

Although Cd is known to be carcinogenic in humans, we intended to utilize its osteotoxic effects by implementing it into bone cement to overcome the local recurrence of OS. A nude mice xenograft study was planned to evaluate the preventive effect of Cd on OS cells. The aim of this study was to show whether local application of Cd-impregnated bone cements can induce apoptosis and decrease the viability rates of residual OS cells in nude mice.

Materials and Methods

Cell culture

The effect of the agents to be used in vivo in the OS nude mouse model were first investigated in vitro, to determine the effect on the K7M2 OS cell line for dose optimizations. For this purpose, a preliminary study was performed. The most effective dose in terms of antitumor effect was found to be approximately 5 μM. However, the 0–10 μM dose range was studied with different and repeated intervals. Therefore, the cells were cultured in DMEM (Dulbecco’s Modified Eagle’s medium) containing 10% fetal bovine serum (1% L-Glutamine and 1% penicillin/streptomycin) and incubated at 37°C under 5% CO₂ conditions. When the cells reached about 90% confluency, they were collected using trypsin-EDTA and were seeded onto a 96-well plate (5,000 cells per well). The cells were treated with different doses of Cd (0, 0.1 μM, 0.25 μM, 0.5 μM, 1 μM, 2 μM, 2.5 μM, 4 μM, 5 μM, 7.5 μM, 10 MM) for 24 hours. The cell viability was determined by WST-1 when the incubation period was over. LD90 doses were optimized for the experimental animal. This dose corresponded to the total dose per gram for the animal.

Determination of cell proliferation with WST-1′

In a 96-well plate containing cells, an empty well was used as a control including only media. For each 100 μL/well containing 10⁶ cells/well, WST1 cell proliferation reagent was added (1:10 dilution). The cells were incubated at 37°C in 5% CO₂ for 4 hours. After shaking on a plate shaker for 1 minute, absorbances were measured by ELISA reader at 420–480 nm. The reference wavelength was more than 600 nm (630 nm). The average absorbance of the control group was considered as 100% viability and the relative absorbances were then calculated.

Formation of mouse xenograft model

In this study, female nude mice obtained from Kobay D.H.L. A.Ş. (Ankara, Turkey) with average age of 8 weeks and a mean weight of 30 g were kept in a special room and special cages with hepafiltration ventilation. Mice were kept at room temperature (20±2 ºC) and in 12-hour light/dark conditions during the study period and sterile water was provided. Before starting the study, mice were monitored to ensure that they adapt to the environment for a week. K7M2 cells (1 × 10⁶ cells/mL) were injected by inserting the injector subcutaneously from the right hip into the femur of each animal. The tumor formation was monitored on a daily basis. When tumor volume reached 150 mm³ (approximately in 10 days), the mice were randomized into five groups including seven animals in each.

Experimental Animal Groups

By selecting the mice randomly, the following five groups each containing seven nude mice were formed.

• Group I: Control tumor group

• Group II: Wide resection group

• Group III: Intralesional tumor resection group

• Group IV: Intralesional tumor resection and subcutaneous insertion of physiological saline solution absorbed cements group

• Group V: Intralesional tumor resection and subcutaneous insertion of Cd-absorbed cements group.

Surgical intervention was performed in the class 2 cabinets while the hepafilter is operating. A combination of ketamine 35 mg/kg and xylazine 5 mg/kg general anesthesia was applied to each animal prior to interventional procedures. Under sterile conditions a longitudinal incision was made over the palpable tumor mass. Tumor resection with 1 cm surgical margins was performed in the wide resection group. In intralesional resection groups, tumor tissue was resected with positive margins aiming to leave 15 mm³ of macroscopic tumor tissue behind. In group 3, the defect was left empty; groups 4 and 5 received bone cements prepared with saline and Cd solutions, respectively. Incisions were sutured with 4/0 Vicryl using the standard primary sutures. Postoperative care was performed in IVC (individually ventilated cage) cages. All animals were allowed to feed ad libitum and move freely in their cages. The mice were sacrificed 15 days after the operation. Ether over-inhalation anesthesia was administered prior to sacrification. The organs and tissues of the mice were dissected. Surgical resection sites were evaluated histopathologically in each group. Any tumor recurrence was investigated in group 2. In groups 3–5 residual tumors were dissected, any progression or regression in tumor sizes was noted and then residual tumor bed was placed in formol to be embedded into paraffin.

Chemicals

Cadmium

A liquid form (Merck) was provided and tested in vitro. One cubic centimeter was mixed with bone cement.

Bone cement

Biomet Bone Cement (Biomet Orthopedics Warsaw, Indiana, USA) was prepared by adding Cd and saline in accordance with the experimental group.

Evaluation of apoptosis, necrosis, and viability

Three different outcome measures were evaluated to understand the effect of Cd on OS cells. Apoptosis, which can be triggered via various physical, chemical, and biological factors, is the most common form of programmed cell death. Apoptosis was assessed with the TUNEL method. However, necrosis is a form of cell injury resulting in the premature death of cells, caused by external factors such as infection, toxins, or trauma. We evaluated necrosis histomorphologically with hematoxylin and eosin (H&E) staining. Viability is the ratio of live cells after the analysis of early and late apoptotic cells, which was evaluated with flow cytometry using annexin V-propidium iodide.

Assessment of apoptosis with the TUNEL in paraffin section

Paraffin-embedded tumor tissue sections were kept at 60ºC in an oven for one night and then kept twice in xylene for 30 minutes. The sections were exposed to 96%, 80%, 70%, and 60% ethyl alcohol for 2 minutes, respectively and washed with PBS (Phosphate-buffered saline) for 5 minutes. The circumference of the sections was drawn with the bounding pen and kept at room temperature for 15 minutes with Proteinase K solution at a dilution of 1: 500. After three times for 5-minute washing with PBS solution, endogenous peroxide was blocked by 3% H₂O₂ exposure (5 minutes). After washing, 100 μL of TdT solution was prepared for each section (77 μL of reaction buffer solution + 33 μL TdT), then dropped onto the sections and incubated at 37°C for 1 hour. Antidigoxigenin conjugate was kept at room temperature for 30 minutes. After washing, the DAB solution was added dropwise and kept in a humidified and dark for 5–10 minutes at room temperature. The cells were stained with Mayer’s hematoxilene for 1–5 minutes. After washing with distilled water, the sections were placed into 80%, 96%, and 100% ethyl alcohol, respectively, for 1 minute each and the sections were dried twice for 5 minutes in the xylene. Sections were covered with a coverslip. Apoptosis staining was performed with the TUNEL method in 5-μm-thick polysine-coated slides prepared from paraffin block. Evaluation was performed by counting in 1.000 different cells in five different fields and recording an average value.

Assessment of necrosis with H&E staining

After routine staining of the fresh tumor tissues with H&E staining procedure, the tissues were evaluated by an experienced pathologist under a microscope. Number and ratio of necrotic cells were identified histomorphologically.

Assessment of viability with flow cytometric apoptosis evaluation using annexin V-propidium iodide staining protocol

Tumor samples were freshly prepared for flow cytometry study in RPMI (Roswell Park Memorial Institute) cell culture medium. Tumors were divided into small pieces in a medium, filtered, and a single-cell suspension was obtained. Cells were centrifuged at 1,200 rpm for 5 minutes and were resuspended in 200 μL of PBS. For each sample, 100 μL of cell suspension was transferred to an unstained tube used for correct gating. The remaining 100 μL was taken into another 5 mL polystyrene flow cytometry tube for annexin V and PI staining. During staining, the tubes were kept in ice and were wrapped with an aluminum foil to protect them from light. Five microliters of annexin V, which is FITC conjugated, and 10 μL of PI were added to the tubes to be stained. Tubes were incubated at +4 ° C for 15 minute in a dark environment. During incubation, 10× annexin V binding buffer was freshly diluted to 1× concentration with PBS. At the end of the incubation, 400 μL of 1× annexin binding buffer was added to all tubes (containing unstained and annexin V PI). Tubes were evaluated with BD Accuri C6 flow cytometry device for apoptosis, necrosis, and viability results. Annexin binds to early apoptotic cells because of the ends of phosphatidylserine, which normally face cytosol, and turns to the extracellular direction with the onset of apoptosis, while PI binds to completely dead cells in late stage of apoptosis. In the dot plot analysis, the double negative part that does not bind to either of these markers indicates the alive cells, which gives us the viability ratio.

Statistical analysis

All data are provided as mean ± standard deviation. The statistical analysis was performed using the Statistical Package for Social Sciences version 15.0 (SPSS Inc.; Chicago, IL, USA) of the Windows statistical program with p<0.05 considered statistically significant. Nonparametric tests, Kruskal–Wallis, and Bonferroni Mann-Whitney U test were used for comparing means between groups.

Ethical approval

This study was approved by the Dokuz Eylül University Laboratory Animal Local Ethical Committee, with the protocol number 41/2014.

Results

Tumor was formed within 1 month after subcutaneous injection. No sepsis or wound problems were observed. Only one animal in the control group lost more than 15% of its weight and was excluded from the study. Ethical committee was informed about the excluded case.

Macroscopic findings

No significant difference was observed macroscopically between groups 1 and 3; residual tumor sizes in group 3 were similar to tumor sizes in control group (Figure 1). Wide resection group also developed palpable recurrent tumor in nude mice; however, these were rather smaller compared to groups 1 and 3. When the bone cement groups were evaluated, it was macroscopically observed that the final size of the residual tumor developed in group 4 was larger compared to group 5 as shown in Figure 2. In contrast, only a thin layer of residual tumor was observed around bone cement in group 5 (Figure 3).

Figure 1.

Appearance of primary tumor in the control group (right); and residual tumor lesion (left) 15 days after intralesional resection

Figure 2.

Appearance of residual tumor lesion in intralesional excision + cement with SF group (T: tumor, C: cement)

Figure 3.

Appearance of residual tumor lesion in intralesional resection + cement with cadmium group. (T: tumor, C: cement). Note the thinner formation of tumor compared to the residual tumor in Figure 2

Apoptosis, necrosis, and viability findings

In group 1, there was no other intervention rather than tumor formation and therefore apoptosis and necrosis percentages and viability were evaluated in primary tumors, solely. In groups 2–5, two samples for each groups were obtained and evaluated. In these groups, samples gathered at the time of first intervention are labeled as “primary” and samples gathered after sacrification are labeled as “recurrence” for group 2 and as “residual” for group 3–5. The apoptosis, necrosis, apoptosis/necrosis (A/N) ratio, and viability values of each group are given as mean ± standard deviation and are presented in Table 1.

Table 1.

Data showing the mean and standard deviation of apoptosis, necrosis, apoptosis/necrosis ratio (a/n), and viability results of each group included in our study

		Apoptosis	Necrosis	A/N	Viability
Control		3.45±2.09	4.17±2.89	1.4±1.39	90.6±2.02
Wide	primary	1.47±1.19	3.1±1.02	0.46±0.29	95.3±1.99
resection	recurrence	0.95±0.52	4.02±0.75	0.33±0.17	94.6±0.42
Intralesional	primary	3.9±3.65	5.12±4.46	1.17±1.13	90.2±7.43
resection	residual	2.03±1.54	2.8±1.56	0.93±0.90	93.8±2.13
Intralesional	primary	0.9±1.37	5.1±3.83	0.13±021	93.3±5.54
+ cement	residual	2.31±2.58	3.76±1.95	0.55±0.76	92.2±4.73
Intralesional	primary	2.78±3.47	3.34±3.03	1.17±1.01	91±6.85
+ Cd cement	residual	4.02±2.91	4.25±1.90	1.09±0.84	87.9±3.96*

^*indicates statistical significance

Histopathological findings

After the tumors of the control group and the other groups were excised, primary tumor sample of group 1, recurrent tumor sample of group 2, and residual tumor samples of groups 3–5 were removed to be embedded into the paraffin block, and paraffin sections were evaluated histomorphologically. As seen in Figure 4; there is a vascularized, aggressive, muscle-invading undifferentiated tumor in the control group. In group 3, where only intralesional resection was performed, both primary and residual tumors were aggressive and invasive in the same way as the control group. Although a wide excision with 1 cm clean margins was performed in the second group, all mice developed recurrent tumors at the end of 2 weeks. In group 4 vascularized residual tumor was seen around the cement. However, remarkable tumor cell necrosis was observed in group 5.

Figure 4. a–i.

Tumor tissue microscopic sections, (a) Control group; the vascularized aggressive, muscle invasive undifferentiated tumor tissue. (b, c) Intralesional excision group, b: primary tumor, c: Residual tumor at the same site after 2 weeks. Both tumors were aggressive and invasive similar to control group. (d, e) Wide excision group. d: primary tumor and e: recurrent tumor. Although wide excision with 1 cm surgical margins was performed recurrence occurred in all seven nude mice. (f, g) Cement with SF applied after intralesional excision group. f: primary tumor. g: residual tumor around bone cement after 2 weeks. (h, i) Cement with Cadmium applied after intralesional excision group. h: primary tumor. I: residual tumor around cement after 2 weeks showing prominent necrosis

In order to evaluate the systemic toxic effects of Cd, tissues from other organs including heart, lung, kidney, liver, brain were taken during sacrification in the Cd-absorbed cement group. Histological evaluation revealed no pathological findings. Local application of Cd did not cause any systemic effect on these organs (Figure 5).

Figure 5. a–d.

Microscopic sections of organs in cement + cadmium group. (a) Liver. (b) Kidney. (c) Lung. (d) Brain. The microscopic examinations showed no pathological changes. Local administration of cadmium did not cause any systemic effects

According to Kruskal–Wallis nonparametric test, no statistical significance was found between groups in terms of apoptosis (p=0.068), necrosis (p=0.772), A/N ratio (p=0.142). There was a statistically significant difference in viability values (p=0.022). Intralesional resection with Cd cement group had the lowest viability results after sacrification.

Discussion

OS is the most common primary malignant bone tumor (20–22%) (14). Before the 1970s, OS was treated with amputation. The survival was short in which 80% of patients died of metastatic disease (15). With the development of induction and adjuvant chemotherapy protocols, surgical techniques, and progression in radiological staging studies, 90–95% of patients are now being treated with limb-salvage procedures with different reconstruction methods. Long-term survival and cure rates of these patients increased to 60–80% in localized (nonmetastatic) diseases (16–18).

As a result of the neo-adjuvant chemotherapy used for the treatment of OS, a significant improvement was observed in the clinical outcomes. However, postoperative recurrences remain a common problem. This recurrence usually occurs within the first 2 years after the surgery, and the likelihood of recurrence decreases with prolonged follow-up (19). This rare situation can be seen within 5 years and this is defined as the late recurrence of OS (20). Therefore, the search for new approaches to prevent recurrence in postoperative patients is important. Based on this fact, in vivo experimental animal studies are required.

Cd is actually a carcinogenic agent known for its chronic exposure. However, many anticarcinogenic agents are known to trigger toxic and long-term carcinogenesis. Cd is eliminated primarily by urine and its half-life (t 1/2) ranges from 10 to 30 years. Cd is also known as an osteotoxin (21). Cd affects normal bone growth; disturbing normal calcium metabolism, potentially via the protein kinase C (PKC) pathway (22). Cd is known to be a calcium channel blocker (23). It has also been shown that Cd displaces and binds to calcium binding sites with a higher affinity in cytoplasmic compartments (24). Several studies have also observed that Cd increases bone resorbtion in rats, calcium release from chick tibias and decreases osteoid tissues and in embryonic chick femur cultures (25–27). Collagen, the major protein component of bone matrix, is produced by osteoblasts modulated by a PKC-mediated pathway. Cd is proven to affect collagen metabolism by disturbing this pathway leading to skeletal findings of Cd toxicity (28).

In previous experimental animal studies on Cd, Waalkes et al. evaluated the chronic carcinogenic and toxic effects of Cd subcutaneous administration in rats (29). Cd is known to induce apoptosis both in vivo and in vitro in many cells, including osteoblasts. Prior in vitro studies have also documented that CD may alter the viability of osteoblast function and induce osteoblast apoptosis, thereby disrupting bone balance. Hu et al. observed in their in vitro study that apoptosis occurs in MG-63 OS cells exposed to Cd for a short time (24 and 48 hours) and they explained the apoptosis mechanism as a result of increasing the oxygen radical, p38 mitogen-activated protein kinase (MAPK), and inactivating the extracellular signal–regulated kinase pathway (ERK) that induces apoptosis (13). In another study on the effect of Cd on bone metabolism in rat OS model, Cd has been reported to affect calcium metabolism through disruption of collagen synthesis and activation of protein kinase C (28).

However, these studies were in vitro studies conducted on OS cell lines, and have not focused on recurrent or residual tumor tissue biology. Also, the systemic effects of Cd on other tissues were not elucidated. In addition, these studies have applied Cd solutions directly on OS cells, however in most clinical situations chemotherapeutic agents require a deliverer for local applications. Polymethyl methacrylate (PMMA) bone cement stands out as the most widely used carrier material in the literature in local studies of different antineoplastic drugs (30). After surgical treatment of benign or locally aggressive tumors such as chondroblastomas, applications such as adjuvant cryotherapy or phenolization have been tried to reduce recurrence (31). However, after extensive resection of malignant tumors such as OS, larger defects occur and these defects are reconstructed with biological methods or with endoprosthetic reconstructions. PMMA cements are widely used for fixing the orthopedic implants to bone during reconstruction surgeries. Therefore, most of the in vivo and in vitro studies have investigated PMMA as a carrier agent. It has been reported that numerous antineoplastic drugs administered in PMMA can be released from bone cement and have an inhibitory effect on tumor cells (32). In the light of these successful literature results, we preferred PMMA bone cement, which we use frequently in our surgical applications, as the carrier material for Cd in our study.

There is a consensus in the literature that drug release from bone cement is very high immediately after administration and tends to decrease over time. For this purpose, some authors aimed to prolong this period by applying agents such as chitosan or mannitol into bone cement. Liu et al. in their study added methotrexate into bone cement, and reported that chitosan prolongs drug release time and provides better integration into bone (33). In another study in which danorubicin-added bone cement was examined on athymic nude mice and Wistar rats, it was shown that the addition of mannitol to the mixture could increase both in vivo and in vitro drug release by up to 90% (34). However, no studies have been conducted to see how these added materials change the biomechanical properties of bone cement.

We investigated the release of Cd from bone cement in different doses in our in vitro preliminary study. The MG-63 and K7M2 OS cell lines were divided into four groups. The first group included application of bone cement that absorbed Cd after cement preparation, the second group of bone cements were prepared by adding Cd into bone cement powder and then mixed with its liquid monomer. In group 3, bone cement without any additions, and group 4 received only Cd solution without bone cement. All agents were applied to the cell lines in increasing numbers. The percentages of viability in cell lines was evaluated using WST-1 method after 24 hours. In the group where the cement was prepared by adding Cd before mixing the bone cement; higher cytotoxic effect was detected than the group that consisted of bone cement that absorbed Cd after cement preparation. When Cd solution without bone cement was applied, 5- and 10-fold dilutions showed the highest antiproliferative effect in the study (25% and 42% viability). In the group where bone cement without any Cd was applied, the highest rate of viability was observed reaching up to 80%. This preliminary study has shown that bone cement either prepared with Cd or absorbed Cd after preparation has cytotoxic effects on OS cell lines (35). In the light of the high release values and cytotoxic effects obtained in the preliminary study, considering that any material to be added to bone cement may decrease the biomechanical properties of the cement, we concluded that there is no need to add any material to increase the release from cement and in our study we examined the local effects by adding only Cd into PMMA. Based on the encouraging results of our preliminary study we planned the present in vivo study.

The local chemotherapeutic effects of other agents in PMMA bone cement including methotrexate and cisplatin have also been investigated in several studies. The release rates, systemic toxic effects, and antitumor activities have been evaluated (6–9). Based on the results of these encouraging studies we developed a nude mice xenograft model and tested the efficacy of Cd-embedded bone cements on the viability of residual OS cells after intralesional resection of the primary tumor. Our findings are consistent with the literature knowledge on the toxic effects of Cd on OS cells and confirm these in vitro findings in an in vivo xenograft model. In addition to these outcomes, Cd did not cause any systemic effects on heart, lung, kidney, liver, brain tissues. Therefore, this study indicates that Cd may help decrease the viability of OS cells without causing major adverse events. Also, Cd might be associated with decreased viability of tumor cells and macroscopically smaller tumor formation in 15 days, although a quantitative volumetric analysis was not performed.

To our knowledge there is no study of Cd-absorbed cement implantation in an experimental animal model, and this is the first study to investigate the effect of Cd embedded in bone cement on OS cells. In this unique study, the effect of local implantation of Cd-absorbed bone cement was tested in terms of antitumorigenic effects in the surgical area of tumor resection.

Despite surgical excision with 1 cm clear margins, wide resection group developed recurrent tumors. Both primary and recurrent tumors after wide resection were aggressive, which may be attributed to the lack of adjuvant or neoadjuvant chemotherapy, considering the immune systems of athymic nude mice included in our study. Although residual tumor tissue is never an issue of debate in sarcoma surgeries (as the main surgical aim is to resect the tumor with adequate clean margins), we developed an animal model with the intralesional resection groups to better demonstrate the toxic effect of Cd on OS cells that have been left intentionally. The encouraging results obtained in our study may shed light for further evaluation of osteotoxic effects of Cd especially after wide resection or amputation models in larger animal models.

There are some major limitations of our study which include the before-mentioned absence of an evaluation on the Cd elution time from bone cement, and a quantitative volumetric analysis between tumor masses at the time of sacrification. Also, we did not evaluate blood samples to assess for the levels of Cd in systemic circulation. Instead we tried to evaluate the toxic effect on other systems including heart, lung, kidney, liver, brain tissues histologically. Intralesional resection does not simulate the clinical situation, but we designed the study to intentionally leave tumor tissue on site and place bone cements on this tissue; to make sure there are OS cells, on which we can investigate the effects of Cd. Another problem about intralesional resection groups is that the resection quality is hard to standardize. Despite not performing a precise volume analysis, we resected the tumor tissue with positive margins aiming to leave 15 mm³ of macroscopic tumor tissue behind in intralesional resection groups. Also, a limitation of our study is the follow-up period. Fifteen days may not be enough to define whether the osteotoxic effects of Cd are permanent or temporary effect. Including another group in our study, which would receive the same intervention as group 5 but would not be sacrificed in 15 days and followed until they die naturally would help to understand whether the OS will recur when the effective Cd dose elution to eradicate OS cell is over. Furthermore, another group of mice in which none of the tumors are resected but Cd-absorbed cement is subcutaneously inserted, would help to demonstrate the real effect of Cd on OS tumor volume. Future studies that would include these groups and also perform only wide resections would contribute a better understanding of the effect of Cd and may be better interpreted to possible clinical applications. Our promising findings regarding the osteotoxic effects of Cd on OS cells should be further explored with longer studies on larger animals focusing on not only residual tumor cells, but also primary OS tissue and recurrent tumors after wide resection.

In summary, we have demonstrated that local application of Cd in bone cement decreases the viability of residual OS cells. Histological evaluation also revealed remarkable necrosis in the group treated with Cd-absorbed cement. Besides, no systemic toxic adverse events or histopathological findings related to Cd were observed.

HIGHLIGHTS.

• Cadmium (Cd) can be applied locally in bone cement.

• Local application of Cd in bone cement decreases the viability of residual osteosarcoma cells.

• No systemic toxic adverse events or histopathological findings related to Cd were observed.