Peri-implant tissue response and biodegradation performance of a Mg–1.0Ca–0.5Sr alloy in rat tibia

Biodegradable magnesium (Mg) alloys combine the advantages of traditional metallic implants and biodegradable polymers, having high strength, low density, and a stiffness ideal for bone fracture fixation. A recently developed Mg–Ca–Sr alloy potentially possesses advantageous characteristics over other Mg alloys, such as slower degradation rates and minimal toxicity. In this study, the biocompatibility of this Mg–Ca–Sr alloy was investigated in a rat pin-placement model. Cylindrical pins were inserted in the proximal tibial metaphyses in pre-drilled holes orthogonal to the tibial axis. Implant and bone morphologies were investigated using μCT at 1, 3, and 6 weeks after implant placement. At the same time points, the surrounding tissue was evaluated using H&E, TRAP and Goldner's trichrome staining. Although gas bubbles were observed around the degrading implant at early time points, the bone remained intact with no evidence of microfracture. Principle findings also include new bone formation in the area of the implant, suggesting that the alloy is a promising candidate for biodegradable orthopedic implants.


Introduction
Research on biodegradable magnesium (Mg) alloys is steadily increasing due to their potential to replace currently used fracture fixation devices [1]. Degradable implants eliminate the need for removal surgeries, which are often required for permanent metallic implants such as titanium, stainless steels, and cobalt-chromium alloys [2,3]. Metallic implants can also corrode or release toxic wear products, leading to hostile local or systemic reactions [2,4]. Although biodegradable polymers, such as poly-lactic acid (PLA) or poly-glycolic acid (PGA), are frequently used clinically, these materials are limited in their applications by their relatively low strength compared to traditional metallic implants [5]. Polymeric devices may leave polymeric debris, which can lead to local inflammatory responses and implant loosening [5][6][7][8]. Another unfortunate consequence of the use of these polymeric materials, especially those based on PLA, is the long degradation time resulting in foreign-body reactions after implantation [5,8,9]. Magnesium, on the other hand, degrades rapidly, has a low density, and is generally considered biocompatible.
Moreover, Mg has stiffness similar to that of bone, potentially reducing the risk of stress shielding effects that can occur with traditional metallic implants [2,10,11]. Mg also has high strength relative to polymeric materials and could thereby support load bearing structures, potentially making it ideal for resorbable implants placed into bone. However, controlling the degradation rate of Mg remains a challenge. This is crucial in implant applications since hydrogen gas evolved during degradation can compromise the mechanical support required for fracture healing processes, while rapid degradation also affects the mechanical integrity of the implant. Although surface treatments and coatings can slow the initial rate of degradation, these processes are often complicated and expensive [12], making alloying a more practical option.
Among various systems previously studied, Mg-Ca alloys have shown attractive responses, including improved degradation behaviour and biocompatibility [13][14][15]. Recently, our research group and others have been exploring the potential to create implants based on Mg-Ca-Sr [16,17]. Mg-Ca-Sr alloys have shown reduced degradation rates compared to cast binary Mg-Ca and Mg-Sr alloys with similar alloying amount [13,18]. In addition, Sr may assist in the bone healing process. Administration of Sr, either as a dietary supplement in the form of salts [19][20][21] or when incorporated into implant material [22][23][24], has shown to promote osteoblast proliferation and improve osseointegration. While preliminary in vitro testing on this ternary Mg-Ca-Sr alloy has confirmed reduced degradation rates compared to binary alloys, and non-cytotoxic responses in osteoblastic cell cultures [16], the in vivo response is unknown. Understanding the biodegradation performance of Mg-Ca-Sr alloys and their effect on surrounding tissue is crucial for future design and optimization. The objective of the current study is to assess the local tissue response when Mg-1.0wt.%Ca-0.5wt.%Sr alloy is implanted into the rat tibia.

Alloy preparation
Cylindrical Mg-1.0wt.%Ca-0.5wt.%Sr alloy pins machined from as-cast material were used as implants [16]. Cast samples were cut using a low-speed diamond saw, followed by machining down to a 0.8 mm diameter smooth surface using a lathe. The cutting and machining processes were performed using oil-based lubricants. Alloy pins were cleaned in soap and ethanol and sterilized in an ethylene oxide gas chamber prior to implantation.

Experimental design
24 adult male Sprague-Dawley rats with a body weight of 250-300 g were randomly divided into five groups: Mg pin at 1 week (n=4), Mg pin at 3 weeks (n=4), Mg pin at 6 weeks (n=6), poly-l-lactic (PLLA) pin at 6 weeks (n=6), and naïve healthy controls (n=4). Because PLLA and Mg were only compared at the final time point, a higher sample number at 6 weeks was used to provide a more robust estimate of the mean and standard deviation of both groups.
Cylindrical Mg alloy pins had a diameter of 0.8 mm (as described above). Biodegradable PLLA pins with a diameter of 1.1 mm were obtained commercially (SmartPin®, Conmed Linvatec, Largo, FL) and used for qualitative comparison. All animal work was approved by the University of Florida Institutional Animal Care and Use Committee and conforms to the AAALAC and NIH ethical guidelines for animal research. In accordance with IACUC-approved humane endpoints, one rat in the PLLA pin group was humanely euthanatized due to unanticipated complications during surgery; these complications were unrelated to the pin. One sample from the Mg-1w group was excluded from μCT analysis and quantitative histological evaluation due to inadequate placement discovered during post-processing. One pin in the Mg-1w group was found to be fractured upon dissection, but the sample was included in the analyses as the time of fracture was not known (placement, post-placement, vs dissection).

Surgical procedure
Rats were anesthetized in an isoflurane sleep box (2-5%), then transferred to a sterile field with isoflurane maintained at 1-2% via mask inhalation. A medial approach to the proximal tibia was made using a combination of sharp and blunt dissection. A transverse pilot hole was made using a 0.8 mm diameter Kirschner wire from medial to lateral in the proximal tibial metaphysis medial to the tibial tuberosity. The biodegradable pin was placed by hand into the pilot hole using a small Jacobs Chuck. Following insertion, the pin was cut at the tibial cortical surface.
The subcutaneous layer was closed with 5-0 PDS, and the skin was closed with skin wound clips; wound clips were removed at 14 days post-operation. Buprenorphine was provided intraoperatively and twice a day for 48 hrs to manage post-operative pain. The animals were allowed to move freely in their cages post operation and were monitored daily for body-condition score [25], signs of edema at the surgical site, and a visual assessment of animal mobility. No adverse events were noted. At the predefined time points listed above, animals were humanely euthanized; then, soft tissue was carefully dissected and bone was fixed in neutral buffered 10% formalin solution for 48 hrs, transferred to gauze soaked with phosphate buffered saline, and frozen at -20°C.

μCT analyses
The disarticulated tibia was placed in a vial and immersed in saline. Samples were scanned using a Skyscan 1172 μCT system (Bruker Corporation, Billerica, MA) at its maximum resolution of  Figure 1. Bone morphology and mineral density was analyzed in the trans-axial plane of the tibia. The volume of interest (VOI) was 8 mm long and centered on the implant. Due to irregular callus formation, two different thresholds were used, one for the highdensity cortical bone and a lower threshold for the lower density bone forming as part of the fracture healing process. Bone density measurement was performed on all samples that received alloy pins, and on controls, by analyzing the entire tissue area (excluding voids). Two hydroxyapatite (HA) phantoms with known density (0.25 and 0.75 gHA/cm 3 ) and similar diameter to the rat tibias were used to convert linear attenuation coefficient to tissue mineral density. One sample from the Mg-1w group was excluded from mineral density measurements as the VOI extended too far proximally. and Goldner's trichrome to study cellularity and mineralization, respectively. Tartrate-resistant acid phosphatase (TRAP) staining was conducted to study bone remodeling. Histologic staining was carried out using standard procedures, with samples qualitatively evaluated using light microscopy (Nikon Eclipse TE2000-U, Nikon Instruments Inc., Melville, NY). The level of TRAP-positive osteoclasts was determined using a computer-assisted histomorphometry approach (MATLAB R2012b, The MathWorks Inc, Natick, MA). An area percentage was determined in each sample by counting red-positive pixels in a radius of approximately 2 mm centered on the implant. For each sample, a minimum of five images (10x magnification) were analyzed from each quadrant around the pin center.

Statistics
Statistical analyses were performed using one-way analysis of variance (ANOVA) with Mann-Whitney post-hoc tests (± = 0.05, Minitab 15, Minitab Inc., State College, PA). A non-parametric method was chosen since the raw data appeared skewed. Mean ± 1 standard deviation is shown in all figures unless otherwise specified.

Initial biological response
All rats tolerated the implants well with only mild wound swelling from surgery (< 4 days) and no visible evidence of infection. No incidences of limb disuse or lameness were noted in any group. Figure 2 shows representative 3D reconstructions of the alloy pin at 1w, 3w, and 6w, respectively, along with the calculated average implant volume. 2D area measurements and calculations of implant cross-sectional area were also performed to complement the 3D calculations since inherent VOI errors exist with 3D reconstructions. The implant was almost completely degraded at 6w with less than 10% of the original volume present, which is in accordance with the 2D calculations (not shown). The remaining implant was located in, or adjacent to, the cortical bone. Within the trabecular region, the implant appeared to degrade uniformly. The PLLA pin was intact and did not resorb during the study period.       At all time points, bone remodeling was observed, as suggested by the osteoblasts (blue/black) and osteoid (red) areas in Figure 6 (Goldner's stain), and the red-colored osteoclasts in Figure 9 (TRAP stain). Mineralized bone is shown in Figure 6, as represented by blue-green color, and newly non-mineralized areas, as represented by red. After 6w, the cortical area at the implantation site was mineralized, showing minimal remnant implant and smooth remodeling of cortical bone (Figure 6). This was observed in all Mg-6w samples, although to different extents.

Degradation performance
Earlier time points had varied amounts of mineralization in the cortical area. At all time points, scattered bone tissue formations were seen in the medullary cavity.

Osteoclasts quantification
The average level of osteoclasts for each group is shown in Table I  Biodegradable Mg alloys attempt to combine the advantages of traditional metallic implants and biodegradable polymers. Mitigating gas production by slowing the degradation rate is one of the objectives of alloying. While it is critical to prevent patient discomfort and maintain implant stability, a potentially fatal effect of rapid degradation can occur if, for example, gas bubbles were to reach the vascular system. In the development process of these materials, in vivo studies provide crucial information on alloy degradation behaviour and biocompatibility. In this study, histological evidence suggests gas formation in the early stages of alloy degradation, which correlates to the decrease in implant volume. More rapid alloy degradation at the early stages is potentially explained by an increased amount of fluid in the cancellous tissue from pin insertion as a result of tissue trauma or fluid accumulation. Faster alloy degradation in the marrow can be attributed to greater vascularization of the cancellous bone, as has been observed and discussed by other researchers [26][27][28]. A richer blood supply could remove degradation products and neutralize the pH more rapidly, which increases the degradation rate since an alkaline environment promotes a protective hydroxide layer that retards further degradation [29,30]. The gas composition was not measured, but according to the electrochemical reaction Mg + 2H 2 O  Mg(OH) 2 + H 2 , the Mg-based pins degrade by giving off hydrogen gas. The hydrogen is then exchanged with the environment [31], diffuses through the tissues and is eliminated by the blood flow and renal system [32]. If diffusion is slower than the gas production, gas will accumulate in tissue cavities. The absence of voids observed at 6w, suggested degradation slowed to allow for complete gas absorption by this time. Similar results have been observed in previous studies on Mg alloys [13,[32][33][34].
The thermodynamic predictions along with previous work in the field suggests that the layer forming on Mg alloys is Mg(OH) 2 [30,35]. For the pins in the current study, it can be hypothesized that the Mg is potentially substituted with small amounts of Ca and Sr. Mg(OH) 2 is considered safe in the body and has been suggested to comprise the osteoconductive properties observed of Mg [36]. The Mg(OH) 2 will react to form soluble MgCl 2 , but its presence could be stabilized by local accumulation of Mg-Ca phosphates [30,34]. The amount of Mg and Ca released from the implant is considered safe [37]. While the Sr content in the tissue was not measured, the amount of Sr released from the pin, even in the event of a burst release, is much smaller than the clinical therapeutic dose for Sr ranelate (5-20 mg per kilo per day) [38].
Bone tissue mineral density appears to increase with time. Since new bone is forming as the pin degrades, an increase in mineral density value with time confirms mineralization of the new tissue took place. It is expected that the mineral density is highest at the final time point, as the tissue had longer time to heal. Large scatter in the data at 1w and 3w is attributed to differences between animals, which are amplified at earlier time points due to trauma from pin insertion and differences in pin degradation.
Histology revealed few inflammatory cells and mild fibrosis, as seen at the early time points. This is expected post-surgery for an inert biomaterial and can be a sign of implant integration [39]. The few accumulations of inflammatory cells that were observed around the alloy pins were located around tissue cavities associated with gas formation, making the inflammatory cells more prevalent at the 1 and 3w time points. Bone resident cell infiltration was observed around all alloy implants while the PLLA was surrounded by fibrous tissue. The fibrous layer separating the bone and implant can be undesirable as it may jeopardize implant stability, which can lead to fixation failure [5,39] and also prevent further implant degradation.
Osteoclast activity represents bone remodeling, which is naturally occurring in bone tissue, but can also be an implication of implant separation [40]. Osteoclasts were primarily located in the cortical region and around tissue cavities. It is unclear if this trend is related to damage and consecutive bone remodeling resulting from drilling and pin insertion or a response to gas generation, or a combination of factors.
The current study investigates the in vivo response to the as-cast Mg-1Ca-0.5Sr alloy, with promising results. It is worth noting, that properties of the material, such as degradation rate, can be tailored by processing, alloying, coatings etc. [16,41,42], which may be relevant for future studies towards clinical applications. In addition to controlled degradation and biocompatibility, adequate strength of the material is important clinically, to avoid failure of the implant. One of the pins in the current study was found fractured upon dissection. Unfortunately the time of fracture is unknown, however it was still used in the data analysis. The alloy's strength has been reported elsewhere [16] by the authors.

Conclusions
This paper examines the in vivo local tissue response to a Mg-1.0wt.%Ca-0.5wt.%Sr alloy for degradable implant orthopedic applications. Despite early gas formation, the rat and local tissue responded reasonably well to the alloy implant with only a mild inflammatory response. Cell infiltration was observed around the degrading implant, with new bone replacing the implant as it degrades, with no histological evidence of fibrous encapsulation. In addition, mineralized bone was seen in the cortical region after 6w, with regained bone mineral density. On the contrary, PLLA biodegradable pins were intact at the end of the study period and were surrounded by fibrous tissue of varying thickness. Overall, there was minimal adverse host reaction, although an increased osteoclast activity around implants was observed at the early stages, which was attributed to tissue healing and bone remodeling. While alloy biocompatibility in the rat was confirmed, the effect of applied stress on the degradation behaviour, and its resultant effect on bone remodeling and strength, has yet to be specifically investigated. Relevant future investigations also include degradation studies determining the effect of implant morphology and location in the body.