Biodegradation of Pure Magnesium and Bone Tissue Reaction in Rabbit Femur 1 Year Results of 3D Micro-CT Monitoring and Histological Observation
2017 Volume 58 Issue 1 Pages 118-122
Details
2017 Volume 58 Issue 1 Pages 118-122
Pure magnesium and its alloys are biocompatible and biodegradable. In cardiovascular surgery, they have been experimentally applied for short-term use as tiny devices. Many studies have been performed on rats and mice using X-ray imaging and CT scanning. However, these small animals have a low radiation resistance and the lethal exposure dosage is small. In orthopedic surgery, fracture fixation using magnesium materials has high potential applicability. Although long-term stable fixation is required, few long-term animal studies have been performed. Therefore, many unclear issues still remain. Accordingly a long-term animal study was performed on Dutch rabbits to investigate the biodegradation of pure magnesium. Specimens implanted into rabbit femurs showed a volume reduction of 40–50% at 52 weeks. Bone resorption was observed in cancellous bone, and new bone formation and direct contact were partially observed. No magnesium hydroxide was observed in the surrounding area.
Pure magnesium and magnesium alloys have a low Young's modulus and unique biodegradable characteristics. They have thus attracted great attention as biodegradable metal materials.1–6)
In cardiovascular surgery and digestive surgery, they have been experimentally applied for short-term use as coronary stents, staplers and suture wires.7,8) They work as small internal devices in soft tissue and are fully absorbed at an early stage. Their short-term biocompatibility and biodegradability have been well investigated by studies on rats and mice, and many practical studies are currently being performed for clinical applications. In addition, the recently developed technology of X-ray μCT scanning has also been utilized in such animal studies. However, these small animals have a low radiation resistance. Thus, frequent measurement to obtain long-term follow-up results in exposure to a lethal dosage.
In orthopedic surgery, fracture fixation devices are ideal application for biodegradable magnesium materials, which may be used as alternatives to biodegradable polymers owing to their mechanical properties, productivity and reliability. However, it has been suggested that these magnesium materials do not yet meet the requirements for fracture fixation. Because their degradation is faster than fracture healing, magnesium devices cannot provide stable fixation for the necessary period. Accordingly, many studies on alloying and their surface coating with biodegradable polymers or calcium phosphate ceramics are now in progress.9–35)
There have been few long-term animal studies on the biodegradation of magnesium materials. Thus, many unclear issues still remain regarding their orthopedic application such as the progress of corrosion, the corresponding bone formation and the soft-tissue reaction.
With this background, an animal study was performed on Dutch rabbits to investigate the long-term biodegradability of pure magnesium implanted into femoral bones, in which the corrosion and volume reduction of the implanted specimen and the bone tissue reaction were periodically examined by 3D μCT. One year after implantation, the femoral bones with the implanted specimens were harvested, and a histological investigation was performed.
The stick-shaped specimens used in this study had a size of 3.0 × 3.2 × 6.0 mm and were made of pure magnesium (99.9%, Rare Metallic Co. Ltd., Tokyo) (Table 1). The surface was polished by #2000 abrasive papers. Prior to implantation, the specimens were sterilized in an autoclave. Consequently the surface oxidation of the specimen was increased, which was confirmed by the high binding energy as observed by X-ray photoelectron spectroscopy (JPS-9200, JEOL Ltd., Tokyo) (Fig. 1).
| Element | Mg | Al | Zn | Mn | Si | Fe | Cu | Ni |
|---|---|---|---|---|---|---|---|---|
| % | Bal. | <0.08 | 0.001 | 0.009 | 0.008 | 0.003 | 0.002 | 0.0003 |
Results of X-ray photoelectron spectroscopy.
Two mature male Dutch rabbits (weight of 1.5–2 kg) were used in this animal study. To minimize the distress to the animals, a general anesthesia was performed by hypodermic injection with a mixture of xylazine and ketamine. In addition, xylocaine was also injected at the surgical site for local pain relief. For the first rabbit, a hole with a diameter of 4 mm was drilled at the lateral femoral condyle. Then the specimen was horizontally implanted into the femoral condyle as a site of cancellous bone (Fig. 2 (a)). For the second rabbit, a hole was drilled at the intercondylar notch in the femur, and the specimen was longitudinally implanted into the intramedullary cavity of the femur as a site of bone marrow (Fig. 2 (b)).
Implantation sites. (a) The Mg specimen was horizontally implanted into the lateral femoral condyle. (b) The Mg specimen was longitudinally implanted into the intramedullary cavity of the femur.
Under sedation with ketamine, CT images of the distal femur were obtained every 4 weeks over 52 weeks using an ultrahigh-speed, high-resolution X-ray 3D μCT system for laboratory animals (R-mCT2, Rigaku Co., Tokyo), which allows the frequent examination of living animals with rapid scanning and a very small radiation dose. The field of view (FOV) ranges from 73 to 5 mm depending on the bore size of the CT gantry, in which the resolving power is fixed to 1/512. Hence, the resolution is improved in a smaller FOV but the examination space becomes smaller. In this study, the FOV was set to 30 mm to ensure a sufficient examination space for the distal femur. Consequently, the exact examination space was ϕ30 × 30 mm, and the voxels were cubes of 59 μm, i.e., the resolution was 59 μm.
The CT data were saved in a storage device in digital imaging and communications in medicine (DICOM) format. Expert INTAGE Ver 1.0 medical image software (Cybernet Systems, Tokyo) was utilized to view 2D and 3D images and for volume measurement. From series of consecutively obtained CT images, the corrosion, the volume reduction of the implanted specimens and the reaction of the surrounding bone tissue were investigated.
For further histological observation, euthanasia was performed with pentobarbital anesthesia at 52 weeks after implantation. The distal femurs with the implanted specimens were harvested and dehydrated with ethyl alcohol solution. Resorcin-fuchsin staining was subsequently performed, in which elastin is stained blue-black, and collagen fiber, such as fibrous connective tissue, the ground substance of vessel walls and the periosteum, is stained pink-red.
Then the stained distal femurs were embedded in polymethyl methacrylate (PMMA) and cut into thin slices of 100 μm thickness. Contact micro-radiograms (CMRs) of the thin specimen were taken using soft X-rays. Finally microscopic observation was performed on both the thin slices and the CMRs.
Because the X-ray absorption rate of pure magnesium is almost the same as that of femoral cortical bone, satisfactory CT images were obtained that clearly show the trabecular pattern of cancellous bone with no metal artifacts.
In the specimen implanted into the femoral condyle, the medial tip was in contact with cancellous bone and the lateral tip was in contact with the soft tissue at the drill hole (Fig. 3). The specimens had a uniform light gray color of medium density with a distinct linear outline. At 4 weeks, the drill hole was closed by newly formed bone. Also, the formation of new bone was observed at the anterior part of the specimen. At the same time, the surrounding low-intensity area had increased in size. At 8 weeks, the specimen still had a smooth outline but was mostly isolated from the bone tissue. At 20 weeks, both the surface and the interior of the specimen had markedly corroded and new bone formation was observed at the rugged surface. At 52 weeks after implantation, the specimen showed no trace of its original shape. There were several large low-intensity spots considered to indicate bone resorption due to hydrogen gas generation, the high concentration of released magnesium ions and the large increase in pH due to the generation of magnesium or hydroxide ions by the corrosion of magnesium in the surrounding cancellous bone.15) In contrast, the bone tissue formed bridges to the specimen at some edges, which is considered to indicate good biocompatibility.
CT images of pure magnesium implanted into femoral condyle.
In the specimen implanted into the intramedullary cavity, the proximal tip was in contact with bone marrow and the distal tip was located in the drill hole (Fig. 4). The specimen showed almost the same changes with time as the specimen implanted into the femoral condyle. At 4 weeks, a large low-density area was observed on the distal side. At 8 weeks, the specimen still had a square outline. At 20 weeks, the corrosion had become significant. At 52 weeks after implantation, the specimen showed considerable corrosion with a large hole inside. Similarly to the specimen implanted into the femoral condyle, the bone tissue formed bridges to the specimen at some parts. In addition, no significant changes were observed in the surrounding cortical bone.
CT images of pure magnesium implanted into intramedullary cavity.
The degradation of the implanted specimens was also clear in 3D images. A slightly rough surface was observed in the extracted volume image even immediately after implantation, while a smooth surface was observed in the normal 3D view. The slight roughness is not due to corrosion but to the effect of volume extraction by graph cut segmentation, which is a well-known and effective method of image segmentation used in medical and biological fields.36,37)
Over time, the surface was clearly corroded, creating small pits, and with increasing time, large defects and fragmentation were observed (Fig. 5, Fig. 6). Also, the volumes of the implanted specimens appeared to decrease almost linearly with the implant time with no significant difference between the implant sites (Fig. 7).
3D images of the specimen in the femoral condyle extracted by image processing.
3D images of the specimen in the intramedullary cavity extracted by image processing.
Volume reduction of pure magnesium implanted into rabbit femurs.
For the specimen implanted into the femoral condyle, the initial volume was 56.7 mm2, which decreased to 47.3 mm2 at 20 weeks and 35.5 mm2 at 52 weeks, a reduction of 37.4%. The linear regression equation was y = −0.40x + 55.9 and the correlation coefficient was r = −0.978.
For the specimen implanted into the intramedullary cavity, the initial volume was 57.9 mm2, which decreased to 42.0 mm2 at 20 weeks and 30.2 mm2 at 52 weeks, a reduction of 48%. The linear regression equation was y = −0.55x + 56.1 and the correlation coefficient was r = −0.975.
Regarding the measurement error, the precision of graph cut segmentation can be practically estimated by comparing the reference volume of 57.6 mm2 calculated from the specimen size and the measured initial volumes of 56.7 and 57.9 mm2 immediately after implantation. The estimated measurement errors were −1.5% and 0.5%, which are negligible compared with the large volume reductions in this study. In addition, mature bone tissue directly in contact with a specimen can be miscounted because its density is similar to that of the specimen. However, graph cut segmentation is an interactive method in which the segmentation is manually performed on several key images before it is automatically performed. Therefore, bone tissue can be excluded by careful manual segmentation in the key images.
3.3 Verification of 2D μCT and histological observationFirst, each 2D μCT image was verified by comparison with a CMR image in which the observation plane of the μCT image was adjusted to be as close as possible to that of the CMR image. The 2D μCT images had a vague outline but clearly demonstrated the state of the implanted specimen and the surrounding bone tissue (Fig. 8, Fig. 9).
Histological examination of the implanted specimen in the femoral condyle by 2D μCT and CMR observation and resorcin-fuchsin staining of the same parts at 52 weeks. (a) Horizontal image, (b) Contact area enlarged, (c) Interface gap enlarged, (d) Horizontal thin slice, (e) Contact area enlarged, (f) Interface gap enlarged, (g) Horizontal thin slice, (h) Contact area enlarged, (i) Interface gap enlarged.
Histological examination of the implanted specimen in the intramedullary cavity by 2D μCT and CMR observation and resorcin-fuchsin staining of the same parts at 52 weeks. (a) Sagittal image, (b) Contact area enlarged, (c) Interface gap enlarged, (d) Sagittal thin slice, (e) Contact area enlarged, (f) Interface gap enlarged, (g) Sagittal thin slice, (h) Contact area enlarged, (i) Interface gap enlarged.
In the CMRs, the corrosion of implanted specimens was clear, revealing an irregular concave surface and round holes. At first glance, empty regions with no bone tissue, which were due to bone resorption, were observed in the area surrounding the implanted specimens. However, mature bone tissue was observed along the concave surface as bone ingrowth, which was accompanied by a narrow gap of approximately 100 μm at the interface. Moreover, in a few parts, bone tissue was directly in contact with the specimen surface.
Although biodegradation will reduce the mechanical strength of the specimen owing to the volume reduction and the formation of surface defects, bone ingrowth into the surface defects may enhance the interlocking of the bone tissue and the implanted specimen.
Upon resorcin-fuchsin staining, the proliferation of fibrous connective tissue and further encapsulation were not observed in the area surrounding the implanted specimens. Similarly to in the CMR observation, a narrow gap and the direct bone contact were observed at the interface. The narrow gap was filled with body fluid, which was replaced by PMMA during the preparative embedding. No chemical deposition was found in this microscopic observation, although magnesium normally produces hydrogen gas and magnesium hydroxide upon its reaction with water. Meanwhile, we confirmed that magnesium hydroxide is insoluble in the alcohol solution used for dehydration and in the PMMA used for embedding.
3.4 Issues to be resolvedFracture fixation devices are generally required to provide stable fixation for over 20 weeks to allow fracture healing with callus formation and bone calcification. Furthermore, in the case of delayed union, a longer period is necessary to ensure fracture healing.38)
According to the results of this animal study, although the progress of biodegradation was linear with time, the implanted specimens largely retained their original stick shape at 8 weeks but significant corrosion was observed at 20 weeks. Therefore, the pure magnesium examined in this animal study does not satisfy the requirements of fracture fixation devices. Accordingly, as previously reported, it is necessary to control the solubility to slow the degradation and also to suppress hydrogen gas generation and the release of magnesium ions to a tolerable level to avoid bone resorption.
Regarding the biodegradation of magnesium materials, several articles have reported an inconsistency between in vivo and in vitro studies, where the degradation in a living body was found to be slower than that in serum, saline solution and simulated body fluid.39) Possible reasons for the slower degradation were also suggested.10,29,40–42) As the body fluid remains almost neutral owing to the bicarbonate buffer system, the released magnesium ions formed insoluble salts with phosphoric acid and carbonic acid on the surface, which reduced the rate of degradation. In addition, the calcium phosphate and calcium carbonate deposited on the surface were also considered to inhibit biodegradation.
The in vivo experiment was carried out in a very complex environment involving various factors that may have affected the rate of biodegradation such as organic and inorganic substances, cell components and their activity, metabolism and homeostasis. For the further development of biodegradable magnesium materials, it is necessary to develop and standardize the experimental protocol for in vitro studies to appropriately simulate in vivo environments. Moreover, a further confirmatory in vivo study is necessary to practically investigate the progress of biodegradation and the reaction of surrounding tissue in living bodies.
The biodegradation of pure magnesium over 52 weeks was investigated in an experiment on Dutch rabbits using a 3D μCT system and by histological observation.
Corrosion due to biodegradation was observed not only at the surface, which proceeded in the depth direction and formed small holes. Moreover, the volume reduction had a linear relationship with the implant time. However, the biodegradation rate of pure magnesium does not meet the requirements for fracture fixation devices. It is necessary to control the rates of degradation and ion release by promising methods such as alloying and surface coating.
In this study, encapsulation by fibrous connective tissue was not observed. Also, bone formation was observed in areas surrounding the specimens together with partial direct contact and bone ingrowth into the surface holes. Moreover, bone resorption was also observed in the surrounding cancellous bone.
In addition, it was confirmed that 3D μCT is useful for the in vivo evaluation of biodegradable magnesium and its alloys with a medium-density, also 3D image processing software is effective for enabling segmentation and volume measurement.
This animal study was authorized by the Animal Study Ethics Board of Aichi Medical University in accordance with International Guiding Principles for Biomedical Research Involving Animals announced by the Council for International Organizations of Medical Sciences 1985 (Ref. No. 2014-69).
