PMMA-Fe₃O₄ for internal mechanical support and magnetic thermal ablation of bone tumors

Abstract

Background: Minimally invasive modalities are of great interest in the field of treating bone tumors. However, providing reliable mechanical support and fast killing of tumor cells to achieve rapid recovery of physical function is still challenging in clinical works.

Methods: A material with two functions, mechanical support and magnetic thermal ablation, was developed from Fe₃O₄ nanoparticles (NPs) distributed in a polymethylmethacrylate (PMMA) bone cement. The mechanical properties and efficiency of magnetic field-induced thermal ablation were systematically and successfully evaluated in vitro and ex vivo. CT images and pathological examination were successfully applied to evaluate therapeutic efficacy with a rabbit bone tumor model. Biosafety evaluation was performed with a rabbit in vivo, and a cytotoxicity test was performed in vitro.

Results: An NP content of 6% Fe₃O₄ (PMMA-6% Fe₃O₄, m_Fe: 0.01 g) gave the most suitable performance for in vivo study. At the 56-day follow-up after treatment, bone tumors were ablated without obvious side effects. The pathological examination and new bone formation in CT images clearly illustrate that the bone tumors were completely eliminated. Correspondingly, after treatment, the tendency of bone tumors toward metastasis significantly decreased. Moreover, with well-designed mechanical properties, PMMA-6%Fe₃O₄ implantation endowed tumor-bearing rabbit legs with excellent bio-mimic bone structure and internal support. Biosafety evaluation did not induce an increase or decrease in the immune response, and major functional parameters were all at normal levels.

Conclusion: We have presented a novel, highly efficient and minimally invasive approach for complete bone tumor regression and bone defect repair by magnetic thermal ablation based on PMMA containing Fe₃O₄ NPs; this approach shows excellent heating ability for rabbit VX2 tibial plateau tumor ablation upon exposure to an alternating magnetic field (AMF) and provides mechanical support for bone repair. The new and powerful dual-function implant is a promising minimally invasive agent for the treatment of bone tumors and has good clinical translation potential.

Introduction

The mortality of bone tumors is exceptionally high, and the prognosis is poor, seriously threatening patient lives ^1-3. A devastating consequence of bone tumors is bone destruction, which causes disastrous skeletal-related events. These intractable clinical problems, such as unbearable bone pain, pathological fractures, spinal cord compression, bone deformity and hypercalcemia, are great challenges to surgeons ^{4, 5}. With the loss of bone support, patients are not able to stand and could become bedridden for an extended time. These complications significantly reduce quality of life and increase mortality ⁶. Surgical resection is the traditional choice for primary bone cancer; however, the extensive trauma imposed by major surgery and often amputation is a serious physical and spiritual pain for patients. Therefore, minimally invasive strategies for treating bone tumors are urgently needed. Recently developed percutaneous vertebroplasty (PVP) and percutaneous kyphoplasty (PKP), which are minimally invasive clinical surgeries, could efficiently solve this critical issue. These minimally invasive techniques have been used to treat pathological fractures and decrease pain. Unfortunately, these strategies aim to increase the survival time and quality of life and are unable to effectively decrease the bone tumor burden ^{5, 7, 8}. Radiotherapy and chemotherapy are commonly used to treat bone tumors, but the bone microenvironment provides a protective alcove for tumor cells to resist clinical radiotherapy and chemotherapy, reducing the curative effect ^9-11.

Hyperthermia, which can be applied with the aim of increasing the temperature of tumor tissues to directly destroy tumor cells, provides another promising treatment modality for patients ^{12, 13}. Additionally, hyperthermia is one of the clinically applied therapeutic modalities for solid tumor regression when combined with radiotherapy or chemotherapy, as demonstrated by several randomized studies ^14-17, including studies of hyperthermia induced by radiofrequency, microwave and high-intensity focused ultrasound ^{18, 19}. Light-induced hyperthermia based on functional nanoparticles (NPs), such as Au, MoS₂ and graphene, is another new hyperthermia modality. When a metal NP is illuminated, part of the intercepted light gets absorbed and ultimately dissipated into heat. This photothermal therapy can also induce tumor cell death, but the limited penetration depth of light severely restricts the application of photothermal therapy to bone tumors ^{20, 21}. Magnetic thermal ablation, which involves magnetic nanoparticles, such as Fe₃O₄ and γ-Fe₂O₃, via the application of a controllable external alternating magnetic field (AMF), provides a promising alternative strategy to achieve tumor ablation ^{12, 13, 22}. In addition to the excellent high heating efficiency, the greatest advantage is the lack of biological barriers, allowing the application of magnetic thermal ablation to both superficial and deeply seated tumors. This capability is based on the specific physical features of the adopted radiofrequency electromagnetic waves ^{13, 23}. Moreover, magnetic NPs can perform different functions when mixed with specific media. This multifunction potential increases the possibilities for developing a novel hyperthermic strategy for treating bone tumors.

Among the greatest limitations of all strategies for treating bone tumors are the inability to quickly repair the bone defect, provide reliable mechanical support to achieve fast physical function recovery and avoid potential damage to some important organs, such as the spinal cord, even if the above methods can induce complete tumor cell death. To address these challenges, we have developed a dual-function injectable material inspired by the bone cement that is widely used for the clinical treatment of spinal compression fractures ^{24, 25}. This formulation is aimed at inducing tumor cell death while quickly filling the bone defect simultaneously to provide internal support for the body weight in a minimally invasive manner by injection. Polymethylmethacrylate (PMMA) bone cement, as an FDA-approved material, has been used as a grouting agent in total joint replacement surgery for more than 50 years ²⁶ and is also widely used in PVP for the treatment of osteoporotic fractures ^{24, 27} because of its excellent characteristics, including its stable solidifying process, mechanical strength and good biocompatibility. However, for bone tumor clinical applications, PMMA bone cement can be used only for mechanical support implantation and cannot kill tumor cells or prevent tumor progression. In additional, PMMA is of limited use in patients suffering from osteoporosis owing to its excessively strong stiffness ²⁸. Moreover, the short window for operation and polymerization heat restrict the utilization and promotion of PMMA in clinical bone tumor treatment ²⁹. Fe₃O₄ NPs have been widely used in MH or as a magnetic resonance imaging contrast agent with good biocompatibility ³⁰. The combination of PMMA bone cement and Fe₃O₄ NPs and investigation the properties of this mixture have gradually attracted interest among researchers. A report described PMMA bone cement containing Fe₃O₄ for thermal ablation treatment of cancer in vitro, but this study did not demonstrate the approach in vivo, especially not its mechanical properties and in situ mimic efficiency with clinical bone tumors ³¹. Some researchers have presented a magnetic thermal ablation approach by distributing Fe₃O₄ NPs inside calcium phosphate cement (CPC) ³². Though the treatment effect in vivo was excellent, the drawbacks include poor injectability, long setting time and the need to improve the degradability of CPC ³³. PMMA bone cement provides good injectability for local injection, a shorter setting time to fix Fe₃O₄ NPs to prevent leakage during the liquid-solid phase transition, faster repair of bone defects to achieve fast recovery, and further non-degradability to ensure that there is no Fe₃O₄ NP leakage over a patient's long lifetime.

Herein, we designed an injectable and highly efficient magnetic material by distributing Fe₃O₄ NPs inside PMMA bone cement for the minimally invasive magnetic ablation of bone tumors and to achieve internal mechanical support. PMMA-Fe₃O₄ in the liquid phase was accurately injected into the bone tumor to fill the bone defect under computed tomography (CT) guidance. After PMMA-Fe₃O₄ transformed to the solid phase, an AMF was applied externally. The thermal ablation that occurred in the vicinity of the magnetic bone cement induced tumor cell coagulative necrosis, and the solid cement simultaneously provided an ideal mechanical support for fast physical function recovery to help relieve pain in the patient as soon as possible. Interestingly, we found that the addition of Fe₃O₄ to PMMA significantly improved the quality of the PMMA bone cement, decreasing the potential for fractures in adjacent bone after intervention, while simultaneously demonstrating good biosafety and biocompatibility. Then, for the first time, we employed the established in situ tibial plateau bone tumor model in New Zealand rabbits via a coaxial puncture needle, which simulated the clinical characteristics of bone tumors to the greatest extent, to evaluate the therapeutic effect and prognosis of PMMA-Fe₃O₄. Most importantly, the combined thermal ablation and mechanical support functions are based on materials that are generally regarded as safe and already clinically utilized. All of the above-mentioned benefits could facilitate further rapid clinical translation and provide an instructive way for surgeons to implement bone tumor treatment.

Results and Discussion

Design and synthesis of PMMA-Fe₃O₄ bone cement

In this work, we successfully prepared magnetic PMMA bone cement (Figure 1A-D) for magnetic thermal ablation to ablate tumors. As shown in Table 1, cements with 0, 3, 6 and 9% Fe₃O₄ were designed to test the heating efficacy and mechanical properties with different iron contents.

Figure 1.

Characterization of PMMA-Fe₃O₄ magnetic compressive materials. (A-D) Morphological photos of PMMA-6%Fe₃O₄: (A) PMMA powder, (B) Fe₃O₄ nanoparticles, (C) MMA monomer, (D) injectable PMMA-6%Fe₃O₄. (E) Low-magnification SEM image of polymerized PMMA. The scale bar is 50 μm. (F) High-magnification SEM image of polymerized PMMA. The scale bar is 20 μm. (G) Low-magnification SEM image of polymerized PMMA-6%Fe₃O₄. The scale bar is 50 μm. (H) High-magnification SEM image of polymerized PMMA-6%Fe₃O₄. The scale bar is 20 μm (red arrow: poly methyl methacrylate, blue arrow: zirconium dioxide particulate, yellow arrow: Fe₃O₄ nanoparticles). (I) Magnetic hystersis loop of polymerized PMMA-6%Fe₃O₄. (J) Energy spectrum of PMMA-6%Fe₃O₄. (K-N) Element mapping of the surface of polymerized PMMA-6%Fe₃O₄ solid implant (K: C, L: O, M: Fe, N: Zr, the scale bar is 2.5 μm).

Table 1.

Composition of PMMA-Fe₃O₄.

Bone Cement	Powder (g)		MMA monomer (mL)
	Fe3O4	PMMA
PMMA	0	2.6	1
PMMA-3%Fe3O4	0.1095	2.6	1
PMMA-6%Fe3O4	0.2260	2.6	1
PMMA-9%Fe3O4	0.3560	2.6	1

Morphology characterization of PMMA-Fe₃O₄ bone cement

SEM images of Fe₃O₄ NPs and PMMA-Fe₃O₄ powders are shown in Figure S1A-D; the size of Fe₃O₄ NP was 120.60 ± 38.95 nm, and the size of PMMA was 30 ± 12.36 μm. It was reported that the transition from a polydomain to a monodomain state of Fe₃O₄ NPs is around 60 nm ³⁴, and the mean size of the applied Fe₃O₄ NPs was 120.60 ± 38.95 nm (Figure S1C). Therefore, the Fe₃O₄ NPs in this study were above the monodomain limit. After mixing, the Fe₃O₄ NPs were uniformly distributed in the PMMA powder, and there was no congregated crystallization or connection between the NPs. SEM images of the PMMA bone cement showed a cobblestone-like material, which is the solid phase of PMMA, that was densely distributed, and the zirconium dioxide particles were closely surrounded by PMMA (Figure 1E-F). Due to the special radiographic functionality of PMMA bone cement, zirconium dioxide particles were added as radiopaque agents to enable CT-guided injection and to obtain precise and objective CT image information. SEM images of the mixture with 6% Fe₃O₄ NPs (Figure 1G-H) show many dense and small dots uniformly and densely distributed in the PMMA matrix, which is white and bright, indicating that Fe₃O₄ NPs successfully integrated into the PMMA matrix without damaging its original structure. It was demonstrated that, given the strength of this stable internal structure, Fe₃O₄ NPs rarely escaped from the tumor tissue or penetrated solidified material, thus preventing Fe₃O₄ NPs from accumulating in sensitive organs and improving the biosafety of PMMA-6%Fe₃O₄ magnetic materials. The results of energy spectrum (Figure 1J) showed the atomic percentage of C, O, Fe and Zr were 48.19%, 39.02%, 5.55% and 7.24%, respectively. The weight percentage of C, O, Fe and Zr were 26.63%, 28.73%, 14.25% and 30.39%, respectively. The element mappings (Figure 1K-N) exhibited a uniform distribution of ferrous elements with carbonaceous and oxygen frameworks, indicating good dispersity of Fe₃O₄ NPs in both the liquid and solid phases. More importantly, a uniform distribution of magnetic NPs means a uniform temperature increase, which guarantees safety during heating. The hysteresis loops (Figure 1I and Figure S1E) of PMMA-6%Fe₃O₄ and Fe₃O₄ NPs were narrow, showing their soft magnetic performance. Under 300 K (26.85 °C), the saturation magnetization values of PMMA-6%Fe₃O₄ and Fe₃O₄ NPs were 8.62 emu/g and 93.99 emu/g, respectively, and the coercive forces of PMMA-6%Fe₃O₄ and Fe₃O₄ NPs were 84.45 Oe and 103.34 Oe, respectively. For these soft magnetic materials, the relatively low coercive force and hysteresis loss properties further demonstrated that PMMA-6%Fe₃O₄ could be heated by AMF due to the hysteresis loss of the material and eddy current losses ³⁵.

Evaluation of injectability of PMMA-Fe₃O₄ bone cement

The injectability of PMMA-Fe₃O₄ is one of the most basic and significant factors in its application. This parameter was initially evaluated to demonstrate that in situ injection of the PMMA-Fe₃O₄ could be performed in a minimally invasive manner. Generally, 70 N is the maximum force that can reasonably be applied to a syringe by healthy human thumbs ³⁶. As shown in Figure 2D, the injectability of PMMA-Fe₃O₄ was more than 90% and increased with increasing Fe₃O₄ NP content, indicating excellent injectability. Even when injected into water, the magnetic bone cement maintained its shape and polymerized into a solid implant (Figure S2A-D). The benign injectability and applicability of PMMA-Fe₃O₄ demonstrated that this type of minimally invasive injectable implant can remain integrated after injection into bone tissues with a full blood supply via CT guidance.

Figure 2.

Physical properties of PMMA-Fe₃O₄ magnetic compressive materials. (A) Polymerization temperature-time curves of PMMA with different Fe₃O₄ nanoparticle contents. (B) Setting time of PMMA with different Fe₃O₄ nanoparticle contents. (C) Polymerization temperature of PMMA with different Fe₃O₄ nanoparticle contents. (D)The injectability of PMMA with different Fe₃O₄ nanoparticle contents. (E) The load-displacement curve and (F) corresponding compressive strength of PMMA with different Fe₃O₄ nanoparticle contents. (G) The load-deflection curve of the three-point bending test, (H) corresponding bending strength and (I) bending modulus of PMMA with different Fe₃O₄ nanoparticle contents. (J) The load-displacement curves of tibial plateau tumor (Tumor), healthy rabbit tibia (Normal), tumor-planted rabbit tibia with implantation of PMMA-6%Fe₃O₄ and tumor-planted rabbit tibia with PMMA-6%Fe₃O₄ under magnetic thermal ablation (PMMA-6%Fe₃O₄-H). (K) Corresponding compressive force and (L) stiffness.

Exotherm during the polymerization of PMMA-Fe₃O₄ bone cement

The setting time reflects the polymerization time, which is important for achieving fast repair of bone defects. It is known that patients with pain caused by a vertebral compression fracture can receive immediate pain relief and fast restoration of physical function as long as the bone cement is polymerized, which generally takes less than twenty minutes.

As shown in Figure 2A-B, the setting time ranged from 5 to 6 min among the bone cements with the various iron contents. Statistical analysis showed that setting time significantly increased with increasing iron content. Nonetheless, the solidification process was rapid in all groups, which indicates that the setting time of this clinically applied bone cement is minimally disrupted by the addition of small amounts of Fe₃O₄. A slight increase in setting time indicates more time in the liquid phase, which is beneficial for effective injection under CT guidance.

As shown in Figure 2A-C, the highest temperatures in the center of the material were 62.1 ± 3.5 °C, 59.7 ± 1.5 °C, 57.6 ± 2.0 °C and 55.2 ± 2.4 °C for PMMA, PMMA-3%Fe₃O₄, PMMA-6%Fe₃O₄, and PMMA-9%Fe₃O₄, respectively. Although the highest polymerization temperature in the center decreased with increasing iron content, this temperature was still greater than 50 °C. Therefore, the influence of a high spontaneous polymerization temperature on tumor cells/tissue needs to be further explored. As a consequence, we used PMMA-6%Fe₃O₄ without heat as a control group for an in vivo study.

Mechanical properties of PMMA-Fe₃O₄ bone cement

As a fast bone repair material, PMMA-Fe₃O₄ has the advantage of a rapid liquid-solid phase transition, providing anticompression and antishear capabilities, which are dominant factors for supporting body weight. Therefore, the International Organization for Standardization published its ISO5833:2002(E) guideline with standardized specifications for the mechanical properties in each clinically applied bone cement; in detail, this guideline indicates that the compressive strength, bending strength and bending modulus must be higher than 70 MPa, 50 MPa and 1800 MPa, respectively. The mechanical properties of the different bone cements changed with the iron content and were less than the specified values in the ISO5833:2002(E) guidelines, indicating that the bone cements were not sufficiently strong to support body weight and cannot be used for in vivo applications. This evaluation is one of the most important steps in selecting a safe and reliable magnetic bone cement to guarantee fast clinical translation to benefit bone tumor patients.

The load-displacement curves from the compression test and load-deflection curve from the three-point bending test in this study are shown in Figure 2E and Figure 2G. From the curves, the compressive strength (Figure 2F), bending strength (Figure 2H) and bending modulus (Figure 2I) were calculated. Ultimately, PMMA-9%Fe₃O₄ was excluded from subsequent experiments because the compressive strength was lower than 70 MPa. Excessive mechanical strength may increase the risk of adjacent vertebral fracture after PVP and PKP ^37-39. Therefore, many researchers tried to add various media, such as hyaluronic acid ⁴⁰, linoleic acid ⁴¹, saline solution ²⁸ and even patients' blood ⁴², to decrease the mechanical strength to make PMMA bone cements more suitable for cancellous bone, especially osteoporotic bone. In summary, as long as the mechanical strength meets the ISO5833:2002(E) guideline, appropriate strength reduction is a safe procedure and is even preferred for clinical applications.

In addition to measuring the mechanical properties of the PMMA-Fe₃O₄ material itself, it is also important to check whether this material can precisely restore the anti-compression capability of a tumor-implanted leg. Additionally, we were curious whether the mechanical properties would differ between the heated and unheated groups after the material was injected into the tumor. The load-displacement curves of the ex vivo tibial plateau compression test are shown in Figure 2J. The compressive forces were 179.0 ± 32.0 N, 310.3 ± 70.5 N, 328.0 ± 23.5 N, and 353.0 ± 73.1 N for the Tumor, PMMA-6%Fe₃O₄, PMMA-6%Fe₃O₄-H and Normal groups, respectively (Figure 2K). The stiffnesses were 231.7 ± 73.4 N/mm, 485.7 ± 145.1 N/mm, 501.3 ± 117.0 N/mm, and 562.7 ± 83.0 N/mm for the Tumor, PMMA-6%Fe₃O₄, PMMA-6%Fe₃O₄-H and Normal groups, respectively (Figure 2L). This result indicates that tumor growth in the tibial plateau significantly worsened the mechanical properties of the bone, making the bone susceptible to compression deformation and fracture. After injection of PMMA-6%Fe₃O₄ and its transformation to a solid, regardless of whether heating was used, the mechanical properties recovered and showed no statistically significant difference from the normal tibial plateau in ex vivo measurements.

In vitro magnetic-thermal-induced thermal efficiency evaluation

It is known that magnetic materials can convert electromagnetic energy into heat because of Brown and Néel relaxation, hysteresis loss or eddy current loss ^{43, 44}. The SAR value for Fe in PMMA-6%Fe₃O₄ was 389 W/g under this AMF. As illustrated in the images in Figure 3A, the color of the saline solution without iron barely changed. As the iron content increased and time passed, the temperature of the saline solution increased remarkably (Figure 3B). The temperature reached 63.6 ± 2.3 °C, 76.3 ± 3.0 °C and 82.9 ± 2.6 °C for 150 μL of PMMA-6%Fe₃O₄ (m_Fe: 0.01 g) at the time points of 60 s, 120 s and 180 s, respectively (p < 0.05). The temperature reached only 34.8 ± 0.6 °C and 60.1 ± 2.2 °C for 150 μL of PMMA and PMMA-3%Fe₃O₄, respectively, after 180 s of heating. For 100 μL, 150 μL and 200 μL of PMMA-6%Fe₃O₄, the temperature increased to 63.6 ± 2.3 °C, 70.1 ± 1.4 °C and 80.5 ± 4.3 °C, respectively, after 60 s of heating (p < 0.05) (Figure 3C). As the load of Fe₃O₄ powder in the bone cement increased, the temperature increased linearly with the AMF, proving significant magnetic field-mediated thermal conversion of the PMMA-Fe₃O₄ bone cement. The results of this experiment precluded the use of PMMA-3%Fe₃O₄ because of its low heating efficiency.

Figure 3.

Therapeutic performance of the PMMA-Fe₃O₄ magnetic compressive materials for magnetic thermal ablation. (A) In vitro thermal images of PMMA with different volumes and Fe₃O₄ nanoparticle contents for varied durations; PMMA without Fe₃O₄ and saline solution were used to treat the control groups. (B) The corresponding temperature-time curve at different masses of Fe₃O₄ and (C) PMMA-6%Fe₃O₄ with different volumes.

In the aforementioned work, the PMMA-9%Fe₃O₄ bone cement was excluded because of its poor mechanical properties (compressive strength was less than 70 MPa), which did not meet the ISO5833:2002(E) guideline. Additionally, compared with PMMA-6%Fe₃O₄ and PMMA-9%Fe₃O₄, the heating efficiency of PMMA-3%Fe₃O₄ cement was lowest, and it was not able to achieve the target temperature for tumor ablation and action duration. PMMA-6%Fe₃O₄ bone cement, with good magnetic thermal performance and appropriate mechanical properties, was the most suitable for in vivo study.

Evaluation of magnetic-thermal-induced temperature distribution and ablation efficiency ex vivo

During magnetic thermal treatment, it is a challenge to limit the therapy area in the tumor and avoid injury to surrounding healthy tissue. Infrared imaging can measure surface temperatures but has failed to monitor temperatures in deep tissue. Although MRI can monitor temperatures in deep areas, a locally high content of magnetic material, such as Fe₃O₄ NPs, will produce defects upon MRI examination, which causes real-time MRI monitoring to fail. Tay ZW, et al. shaped magnetic fields to "gate" the heating only to selected regions under MRI guidance ⁴⁵. Jessica F. Liu, et al. developed a magnetic device containing a sharp zero point, allowing for magnetically driven spatially targeted drug release from thermally sensitive liposomal drug carriers ⁴⁶. Therefore, it is highly necessary to take actions such as performing experiments to evaluate heating and ablation areas. We placed the bone cement in excised bovine liver and heated it to simulate heating bone tumors in vivo. The goal of this experiment was to clarify the relationship between heating time and ablation area and further control the ablation area to cover all tumors and prevent extensive injury of the surrounding healthy tissues.

As shown in Figure S4A, thermal images of a 2 cm × 2 cm × 4 cm section of excised bovine liver containing hemispheric PMMA-6%Fe₃O₄ (75 μL, m_Fe: 0.005 g) were obtained to simulate the evaluation of the change in heating temperature in deep tissue; the temperature increased rapidly in the closed tissue and decreased obviously with increasing distance (Figure S4C). The temperature decreased from 82.7 ± 13.8 °C (1 mm away from the surface of the embedded PMMA-6%Fe₃O₄) to 71.2 ± 5.5 °C, 61.4 ± 2.8 °C, 53.8 ± 0.4 °C, 47.5 ± 0.2 °C, 44.5 ± 0.3 °C, 40.8 ± 0.2 °C, 38.4 ± 0.5 °C, and 36.9 ± 0.9 °C (1.5, 2, 2.5, 3, 3.5, 4, 4.5, and 5 mm, respectively, away from the surface of the embedded PMMA-6%Fe₃O₄) after 150 s of heating (Figure S4C). This result indicates that high temperatures can be used to cover all areas of PMMA-6%Fe₃O₄ bone cement while avoiding high temperature disruptions of large areas of the surrounding healthy tissue.

Although the area subjected to high temperature can be controlled, the real ablation distance must be checked. As Figure S4B shows, the thermal images acquired from the surface of the excised bovine liver embedded with 150 μL of PMMA-6%Fe₃O₄ (m_Fe: 0.01 g) showed that the temperature increased with time, and the surface temperature reached only 40 ± 1.3 °C after heating for 180 s, which makes it easy to conclude that the surface temperature does not reflect the deep temperature (Figure S4D). Segmenting the heated liver in half and removing the bone cement showed that the ablated liver tissue became white, and the ablation distance could be measured easily (Figure S4F). The corresponding visually measured ablation distance is shown in Figure S4E. The ablated tissue distances were 1.38 ± 0.12 mm, 3.06 ± 0.24 mm, and 4.82 ± 0.23 mm for 120 s, 150 s and 180 s, respectively. The pathological examination of hematoxylin-eosin (H&E)-stained heated excised bovine liver further demonstrated that the ablation efficiency was exact and controllable. Prussian blue staining demonstrated a lack of Fe₃O₄ particulate leakage during heating, further illustrating that this magnetic bone cement is stable during heating (Figure S4G).

It has been reported that irreversible cellular damage occurs when heating tissues at 43 °C for several hours, and cell death arises with continuous heat for 4-6 min at 50-55 °C. When the temperature reaches 60-100 °C, tissue coagulation occurs nearly immediately, while at temperatures greater than 100 °C, tissue becomes vaporized and carbonized ^47-50. Our results are consistent with the literature, and our target temperature is greater than 60 °C, which can effectively and quickly induce tumor necrosis and stimulate the immune system through the expression of heat shock proteins (HSPs). In summary, according to the above two experiments, the distance that was heated to temperatures greater than 60 °C was limited to 2 mm, the distance heated to temperatures greater than 47 °C was limited to 3 mm, and the ablation distance was limited to 3 mm, preventing damage to normal tissues while enabling ablation of tumors.

Biosafety of PMMA-6%Fe₃O₄ in vivo and in vitro

PMMA bone cement is a widely used material for bone filling and repair in orthopedics that exhibits excellent biocompatibility and biosafety ²⁶, Additionally, Fe₃O₄ is used as a contrast agent in MRI. Therefore, the mixed material should exhibit comparable biocompatibility and safety. Based on the cell apoptosis results measured by flow cytometry (Figure S5), the proportion of living cells was 94.02 ± 0.80% in the control group and 92.04 ± 0.74% in the experimental group. There was no statistically significant difference between the two groups (p > 0.05). The cell experiment indicated that PMMA-6%Fe₃O₄ had no obvious toxic effects. As shown in Figure 4A, there were no inflammatory cells and no obvious damage in the muscle tissues around the PMMA-6%Fe₃O₄, which suggests good biocompatibility. Compared to the corresponding tissues in the normal rabbit, the heart, liver, spleen, lung and kidney tissue showed no changes, and no black Fe₃O₄ particles were found in these viscera, which indicates a lack of organic injury after implantation of this bone cement and an inability of Fe₃O₄ NPs to escape from the bone cement (Figure S6). The serum tests for heart function, liver function and kidney function, including ALT, AST, CR, BUN, CK and LDH assessments, revealed no significant difference at each test point (Figure 4B). The blood tests included WBC, RBC, HB and PLT counts, which were in the normal range (Table S1). These results indicate no effects on heart function, liver function, kidney function and blood function, illustrating the good biosafety of locally injected PMMA-6%Fe₃O₄ in vivo. These results clearly demonstrate the excellent biocompatibility and biosafety of magnetic bone cement.

Figure 4.

Biocompatibility and biosafety of PMMA-6%Fe₃O₄. (A) H&E staining of primary organs (heart, liver, spleen, lung, kidney) and muscle in healthy rabbits with and without injection of PMMA-6%Fe₃O₄ after 4 weeks. The scale bar is 50 μm. (B) Histograms depicting variations in serum biochemical parameters for alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatinine (CR), blood urine nitrogen (BUN), creatine kinase (CK) and lactate dehydrogenase level (LDH-L) containing PMMA-6%Fe₃O₄ with different time points over 28 days.

With the application of inductively coupled plasma optical emission spectrometer (ICP-OES) quantitative measurement, the iron concentrations of PMMA-6%Fe₃O₄ in the first batch with and without heating were 59.85 ± 0.29 mg/g and 60.18 ± 0.40 mg/g, respectively, indicating no significant difference between the two groups (p > 0.05). Furthermore, the Prussian blue staining results for the organs were negative, illustrating that no Fe₃O₄ NPs escaped from the PMMA bone cement after heating (Figure S7). The iron concentrations of PMMA-6%Fe₃O₄ in the second batch with and without attraction by the Nd₂Fe₁₄B magnet were 59.90 ± 0.20 mg/g and 60.03 ± 0.18 mg/g, respectively, showing no significant difference between the two treatments. These results strongly illustrated that no Fe₃O₄ NPs escaped after magnetic heating in vivo and even exposure in a strong magnetic field.

Preparation of a tumor model in the rabbit tibial plateau with a VX2 tumor mass

Generally, nude mouse models with flank xenograft tumors are widely used for magnetic thermal ablation experiments ^{35, 51, 52}, but these models cannot simulate the real change in bone, which may decrease the reliability of therapeutic effects and interfere with fast clinical translation. The basic signs and characteristics of a bone tumor include destruction of bone, formation of tumorous bone, reaction of the periosteum, tumefaction of soft tissue and a change in neighboring tissue ⁵³. Therefore, it is important to establish an ideal bone tumor model to simulate bone tumor invasion and verify the efficacy of magnetic thermal ablation. The rabbit tibial plateau tumor model is an alternative orthotopic tumor model due to its size, which is large enough to inject PMMA-Fe₃O₄ bone cement, and because the tibial plateau is a load-bearing bone; these characteristics can simulate bone tumor invasion in humans. This model can also provide sufficient sample to test both the efficacy of bone tumor ablation and the repair ability of these dual-function materials. The VX2 tumor is a virus-induced skin papilloma of rabbits that was first characterized in 1933 by Shope and Hurst ⁵⁴, and the VX2 tumor model is considered the most suitable model for studying skeletal tumors ^{55, 56}.

Standard methods to establish a rabbit tibial plateau bone tumor with VX2 mass/cells involve cutting the skin to expose the bone surface, using a burr to drill a hole in the tibial plateau, placing a tumor mass in the hole, sealing the hole via bone wax and suturing the skin incision ^{57, 58}. These traditional procedures are complex, which may lower the efficiency of establishing the model. More importantly, in addition to the loss, the extensive trauma caused by drilling bone tissue will easily connect the drilled hole to the bone marrow cavity, which may change the model position and result in tumor metastasis, easily increasing the mortality ^{58, 59}. Therefore, it is critical to modify the methods to develop a novel model. In our method, the diameter of the puncture needle was only 1.2 mm (Figure S8). With the guidance of the needle, the tumor mass could be accurately seeded in the tibial plateau, the operation procedures were extremely convenient and resulted in a small needle tract, and there was no incision to suture. After 13 to 15 days, the bone tumor reached a volume of 180.0 ± 15.0 mm³ and was ideally located in the central cancellous bone of the tibial plateau, which was observed through CT and MRI examinations. Compared with subcutaneously implanted tumor models in mice, the in situ tibial plateau bone tumor model in rabbits better imitates the clinically common tibial plateau bone tumor. Taking advantage of this novel bone tumor model, we obtained a detailed understanding of the progression of bone tumors and demonstrated the exact treatment effect.

In vivo magnetic-thermal-induced ablation efficiency

After injection, the tumor was simply heated by putting the tumor-bearing rabbit leg in a water-cooled magnetic induction coil. Unlike the heating of subcutaneous tumor xenografts in nude mice ^{35, 51}, the skin temperature increased slowly and gradually. Even after continuous heating for 150 s, the peak temperature was less than 55 °C (Figure 5A and Figure 5C). This slow heating process indicated that the heating efficiency could be well controlled, thereby protecting healthy tissues. After heating, there was no obvious damage to the leg, such as skin shrinking, paleness, limb swelling or knee joint activity limitations; only slight transient hyperemia of the leg skin was found. Comparatively, there was no obvious temperature increase in the Tumor-H group after exposure to AMF for 150 s (Figure 5A and Figure 5C).

Figure 5.

In vivo magnetic thermal therapy for tumor and follow-up CT coronal reconstruction examination. (A) Thermal images of rabbit leg in the PMMA-6%Fe₃O₄-H group and Tumor-H group and (C) the corresponding temperature-time curve. (B) Enhanced MRI images and coronal reconstructed CT images at each follow-up time point (red arrow: bone destruction and swelling of soft tissue, blue arrow: cortical bone of upper tibial plateau, yellow arrow: area of bone resorption and new bone formation). (D) Curve of the circumference of the proximal crus change over time. (E) Curve of the CT value of the upper tibial plateau cortical bone change over time. (F) Curve of the thickness of the upper tibial plateau cortical bone change over time.

The therapeutic effect was continuously and regularly examined via CT imaging. As Figure 5B shows, before injection, enhanced MRI images showed a low signal in the central tibial plateau with peripheral enhancement, and the corresponding coronal reconstructed CT images showed that the cancellous bone of the tibial plateau was destroyed by the tumor, leaving a cavity in the bone (Figure S9E) PMMA-6%Fe₃O₄ bone cement was easily be injected into the cavity with CT guidance due to its high fluidity. Postinjection CT images show that the PMMA-6%Fe₃O₄ bone cement distributed and filled the cavity well. This finding indicated that the PMMA-6%Fe₃O₄ bone cement was a highly injectable material and could be precisely injected in vivo under CT guidance. For the Tumor group, there was no obvious change in the CT images after injecting a saline solution.

During the follow-up, the rabbits in the PMMA-6%Fe₃O₄, Tumor and Tumor-H groups showed progressive knee joint swelling, fast growth of soft masses, progressive aggravation of knee joint activity limitations and severe body weight loss. Eventually, these rabbits were unable to stand properly and had to remain prostrate on the ground and depend only on the forelimb to move (Figure S10). Unfortunately, these rabbits were too weak to survive for more than 1 month (Figure 7D) and died for visceral metastasis (Figure S11). In comparison, most rabbits in the PMMA-6%Fe₃O₄-H group showed typical growth for 56 days, except three rabbits that did not survive the treatment duration (Figure 7D). After dissection, two rabbits were found to have visceral metastasis, and the other one died from diarrhea. As shown in Figure 5B and Figure 6A, the volume of bone defects in the PMMA-6%Fe₃O₄-H group increased slightly at the 35-day follow-up. The absorbed bone tissue was located close to the bone marrow, likely caused by heating. The most exciting finding is that new bone formation was found after day 42; this new bone was located in the absorbed area. The CT value and thickness of the upper tibial plateau cortical bone obviously increased in the CT images, and compared to the preinjection measurements, the corresponding measurements on day 56 showed a significant increase (Figure 5E-F). The new bone formation after tumor ablation confirms the excellent controllability of the heating area. PMMA-6%Fe₃O₄ with AMF can completely eliminate bone tumors without creating large lesions in the surrounding healthy tissues. The therapeutic process may result in minor injury of healthy bone structures, and these tissues can be regenerated with time. After injection of PMMA-Fe₃O₄, no tibial plateau fracture occurred in the PMMA-6%Fe₃O₄-H group, and the mobility was not compromised, which means that this dual-function material provides suitable mechanical properties.

Figure 7.

In vivo effects of PMMA-6%Fe₃O₄ on tumor suppression with magnetic thermal therapy. (A) H&E staining on day 1 (scale bar is 50 μm), and H&E staining, TUNEL assay and PCNA assay on day 4 (scale bar is 100 μm) (red dotted line: edge of ablation, blue dotted line: edge of the tumor, black dotted line: edge of removed PMMA-6%Fe₃O₄). (B) and (C) are the apoptosis index (AI) and the proliferation index (PI) of each group. (D) Morbility-free survival curves of tumor-bearing rabbits after different treatments. (E) H&E staining (scale bar is 200 μm) of metastasis organs (heart, liver, spleen, lung, kidney and diaphragm). (F) Visceral metastasis rate of tumor-bearing rabbits after different treatments. (G) HSP-70 levels of rabbit serum in different groups before and after magnetic thermal ablation. (H) IL-2 levels of rabbit serum in different groups before and after magnetic thermal ablation.

Figure 6.

Follow-up CT 3D reconstruction examination. (A) 3D-reconstructed CT images at each follow-up time point. (B) Curve of the volume of bone defect change over time. (C) Curve of the body weight change over time.

In contrast, during the 28-day follow-up, the bone defect volume successively and significantly increased in the PMMA-6%Fe₃O₄, Tumor and Tumor-H groups, as indicated by the comparison between the preinjection and postinjection groups (175.7 ± 7.8 mm³ vs. 412.1 ± 38.6 mm³, 174.3 ± 6.6 mm³ vs. 421.0 ± 30.0 mm³ and 179.9 ± 5.9 mm³ vs. 460.5 ± 28.5 mm³, respectively, p < 0.001, Figure 6B). As the CT images show, both bone and soft tissues around the tibial plateau in the Tumor, Tumor-H and PMMA-6%Fe₃O₄ groups became obviously swollen with time, and the bone of the tibial plateau disappeared, demonstrating massive bone loss (Figure 6A-B). The soft tissues around the tibial plateau also swelled severely in response to the tumor invasion (Figure 5B), so the circumference of the proximal crus increased rapidly (Figure 5D). At the 28-day of follow-up, the weight was significantly higher in the PMMA-6%Fe₃O₄-H group (2.4 ± 0.1 kg) than in the PMMA-6%Fe₃O₄, Tumor and Tumor-H groups (1.4 ± 0.3 kg, 1.7 ± 0.2 kg and 1.5 ± 0.1 kg, respectively, p < 0.001, Figure 6C). These results indicate that normal growth occurred in the PMMA-6%Fe₃O₄-H group, while the weight decreased in the control groups.

The microscopic therapeutic outcomes of tumor cells were evaluated by standard H&E staining. As Figure 7A shows, on day 1 after the intervention, a large area of uniformly red-stained cytoplasm, indicating disordered tissue and marked destruction of cells, was found in the PMMA-6%Fe₃O₄-H group. Comparatively, there was no obvious cell destruction in the control groups, which exhibited normal nuclei and cytoplasm. On day 4, the apoptosis and proliferation inhibition of the tumor cells in the PMMA-6%Fe₃O₄-H group increased, as indicated by TUNEL and PCNA assays, respectively. Comparatively, no obvious apoptosis or proliferation inhibition was observed in the Tumor and Tumor-H groups (Figure 7B-C). In the PMMA-6%Fe₃O₄ group, there was no apoptosis, but slight proliferation inhibition was observed. This proliferation inhibition may result from the temporary high temperature during the polymerization of PMMA and a change in the internal microenvironment of the tumor after PMMA-Fe₃O₄ injection. Pathological analysis clearly demonstrated the therapeutic efficacy of PMMA-Fe₃O₄ implants exposed to an AMF. The morbidity-free survival curve is shown in Figure 7D. As shown in Figure 7F, the visceral metastasis percentages in the PMMA-6%Fe₃O₄-H, PMMA-6%Fe₃O₄, Tumor and Tumor-H groups were 14.3%, 71.4%, 57.1% and 64.3%, respectively. The corresponding visceral pathology indicated by H&E staining is shown in Figure 7E. Notably, two rabbits with visceral metastasis remained in the PMMA-6%Fe₃O₄-H group because the local tumors were ablated. The reason might be that visceral metastasis occurred before magnetic thermal ablation was applied. Like other types of thermal ablation, magnetic thermal ablation still has the shortcoming of incomplete ablation. For identified local tumors, magnetic thermal ablation exhibits high efficiency; however, for unknown metastatic tumors, another therapeutic modality needs to be explored in the future.

It has been reported that magnetic hyperthermia can induce the regression of distant metastatic tumors without heat when another local tumor is exposed to heat ⁶⁰. The HSPs presented in thermal-induced dying tumor cells are responsible for subsequent induction of antitumor immune responses ⁶¹. The importance of HSPs, such as HSP-70, HSP-90 and gp96, in immune reactions has been demonstrated by several reports ^62-64. In this study, HSP-70 was detected 1 day after heating, and HSP-70 levels successively increased over the next 14 days in the heating group. The other two control groups showed no HSP-70 increase. Another significant cytokine is IL-2, which can reduce the immune response and inhibit the immune reaction of activated lymphocytes ⁶⁵. In this study, the expression of IL-2 in the PMMA-6%Fe₃O₄-H group decreased with time and eventually became lower than that in the groups without heating. Overall, the increase in HSP-70 and decrease in IL-2 indicate that hyperthermia can improve the host immune state and enhance antitumor treatments. In the PMMA-6%Fe₃O₄-H group, the level of HSP-70 successively increased, while the level of IL-2 successively decreased during the 14-day follow-up period. Comparatively, there were no obvious changes in the control groups (Figure 7G-H). This result means hyperthermia therapy activated immunomodulation. However, this finding requires further experiments to evaluate the antitumor potential.

Conclusions

We have presented a highly efficient and minimally invasive approach for complete bone tumor regression and bone defect repair by magnetic thermal ablation based on PMMA containing Fe₃O₄ NPs; this approach shows excellent heating ability for rabbit VX2 tibial plateau tumor ablation upon exposure to an AMF and provides mechanical support for bone repair. The new and powerful dual-function implant is a promising minimally invasive agent for the treatment of bone tumors and has good clinical translation potential.

Supplementary Material

Supplementary materials and methods, figures and table.

Abbreviations

NPs

nanoparticles

PMMA

polymethylmethacrylate

SEM

scanning electron microscopy

Tmax

maximum temperature

Tset

setting time

CT

computed tomography

AMF

alternating magnetic field

PVP

percutaneous vertebroplasty

PKP

percutaneous kyphoplasty

CPC

calcium phosphate cement

MRI

magnetic resonance imaging

H&E

hematoxylin-eosin

HSPs

heat shock proteins

ICP-OES

inductively coupled plasma optical emission spectrometer

UVECs

umbilical vein endothelial cells

TUNEL

terminal deoxynucleotidyl transferase dUTP nick end labeling

PCNA

proliferating cell nuclear antigen

PI

proliferation index

AI

apoptotic index

IL-2

interleukin 2

WBC

white blood cell

RBC

red blood cell

HB

hemoglobin

PLT

platelet

ALT

alanine aminotransferase

AST

aspartate aminotransferase

CR

creatinine

BUN

blood urea nitrogen

CK

creatine kinase

LDH-L

lactate dehydrogenase levels