In vitro and in vivo assessment of biomedical Mg–Ca alloys for bone implant applications

Hydrogen evolution and pH change

Each set of Mg–0.5Ca and Mg–5.0Ca samples was maintained in 50 mm graduated tubes filled with PBS (Amresco, Korea, pH 7.4) solution and stored upside down to allow any gas that evolved to remain at the top. These tubes were clamped in a PBS jar to protect them from air entrainment and placed at 37°C. The solution level inside the tube was measured intermittently for 24 h. Gas evolution measurements were recorded for each sample in triplicate. Each sample set was kept inside a 15 ml tube containing PBS solution. The pH of the solution was measured with a pH meter (Thermo Scientific, Korea) for 72 h. PBS solutions were replaced by fresh ones after 12 h during the whole immersion period. A mean of three readings was considered for each sample.

Immersion test

To determine material stability of the Mg–Ca alloys in an aqueous environment, samples were immersed in PBS solution for 24 h. Mg–0.5Ca and Mg–5.0Ca samples were immersed in 20 ml of PBS solution. The temperature was maintained at 37±0.5°C during the test. The immersed samples were washed gently with deionized water, and air dried prior to further characterization. The change in surface morphology, chemical composition, and phase analysis of each sample after degradation was evaluated by SEM, EDX, and XRD, respectively.

Implantation of Mg alloy into rabbit femur

New Zealand white rabbits were employed for in vivo implantation of Mg–0.5Ca and Mg–5.0 Ca samples. The protocol for the care and use of laboratory animals was approved by the Animal Ethical Committee of Soonchunhyang University, South Korea and duly observed (SCH 16-0014). Three rabbits were used for each sample. Two time intervals (2 weeks and 4 weeks after implantation) were chosen for in vivo studies. Rabbits (weight: 3 kg) were housed individually in stainless-steel cages. Anesthesia was induced in rabbits at 5.0% isoflurane and oxygen flow rate of 800 ml/min using an induction chamber connected to an isoflurane vaporizer (Harvard apparatus, USA). Upon complete loss of consciousness and righting reflex, the surgical site along the lateral area of the rabbit’s right knee was shaved, cleaned with 70% ethanol, and disinfected with povidone-iodine solution. A 3–4 cm full thickness skin incision was made along the lateral surface of the knee joint to expose the underlying joint tissues. A small incision measuring 1–1.5 cm was made on the exposed ligament to reveal the lateral epicondyle of the distal femoral epiphysis. Using an 8.0 mm diameter trephine drill, an 8.0mm diameter by 3.0 mm deep defect was created on the femoral condyle. Similar sized Mg alloy plates containing either 0.5 wt% Ca or 5.0 wt% Ca were then placed in individual the defects. Surgical incision on ligament and skin was then repositioned and closed using sutures. Tramadol was given orally for pain management and Baytril antibiotic (Bayer Korea) was injected intramuscularly as a prophylactic. After 2 and 4 weeks, rabbits were sacrificed via isoflurane induction and air embolism. Implanted Mg alloy plates were then extracted from the surgical sites and fixed in 10% formalin prior to micro-CT scanning and histological preparation.

Micro-CT and histological analysis

Extracted tissue samples were scanned with Skyscan 1076 micro-CT (Bruker micro-CT, Belgium) scanner to generate relevant datasets containing mapped-out tissue densities. Micro-CT datasets were then reconstructed using NRecon software (Bruker micro-CT), after which the resulting reconstructed data was analyzed for bone volume/tissue volume (BV/TV%) in CTAn software (Bruker micro-CT) using a uniform volume of interest. Three-dimensional analysis of the analyzed data was then rendered using CTVox. Formalin-fixed tissue samples containing circular Mg implant were transversally cut along the axis of the implant using a diamond abrasive cutter (Topmet; R&B, Seoul, Korea). The cross-sectional area of the implant site was then exposed for closer observation. Micrographs of the exposed implant site were taken using a hand-held digital microscope (Dino-Lite pro; AnMo Electronics, Taiwan, New Taipei City). Tissue samples were then decalcified and dehydrated using alcohol series followed by xylene immersion. The xylene-treated samples were then immersed in molten paraffin wax prior to embedding. Tissue sections measuring 5.0 µm thick were then cut with a microtome (Thermo-Scientific, USA) from the blocks and attached on glass slides. The tissue slides were then deparaffinized and re-hydrated with xylene and an alcohol series for 3 min each, respectively. The slides were then washed with distilled water and finally stained with hematoxylin and eosin (H&E). Tissue sections were viewed using an optical microscope (BX53 Olympus, USA) equipped with a DP72 digital camera.

pH change and hydrogen evolution

To analyze the detailed in vitro degradation behavior of alloys, hydrogen gas evolution and pH change behavior of Mg–xCa alloys with different Ca contents (0.5, 1.6, 3.8, 5.0 wt%) as a function of immersion time in PBS solution were determined. Results are shown in Figure 3. It was observed that the hydrogen evolution rate was increased when Ca content in the alloy was increased. The volume of released hydrogen was the least in the Mg–0.5Ca alloy compared with that in others (Figure 3(a)). It is known that gas evolution behavior is directly related to corrosion trend. ⁷ These results suggest that Mg–0.5Ca alloys are less prone to corrosion attack than alloys with other given compositions.

OPEN IN VIEWER

Figure 3.

(a) Hydrogen evolution and (b) pH change behavior of varied Mg–Ca alloys as a function of immersion time in PBS solution at 37°C.

The pH value of a solution plays a crucial role during the immersion test. It has been well established that a rise in pH value is unsuitable for effective cell growth or cell proliferation. It can result in hemolysis of around 66%.^4,23 Moreover, the pH of the solution is also directly related to the corrosion rate of the alloy.^3,24 The results of pH changes in PBS after immersion of different alloy compositions (0.5, 1.6, 3.8, 5.0 wt% Ca) over 72 h are shown in Figure 3(b). It was clearly seen that the pH value of the solution was drastically increased initially with increased immersion time for all samples. Such initial pH rise for all alloy compositions was relatively rapid except for the 0.5 wt% Ca alloy (Figure 3(b)). In particular, pH value was rapidly increased to more than 12 after 3 h of immersion for alloy samples with high Ca contents (3.4 and 5 wt% Ca) and seemed almost stable afterwards. The pH of 1.6 wt% Ca alloy reached around 11.1 during the first 3 h of immersion. It continued to rise until 12 h, although this increase (up to pH 12) was less dramatic compared with the rise in the initial 3 h of immersion. On the other hand, the 0.5 wt% Ca alloy showed gradual increase in pH for the first 24 h (up to 11.2). It then slowly reached its equilibrium pH after 72 h of immersion. The initial spike in pH value after immersion was ascribed to accumulation of Mg(OH)₂ on the surfaces of all alloy samples. However, the pH value stabilized after a longer immersion time.

It was observed that the pH value was higher when the Ca content in the alloy was higher. Higher Ca content influenced pH change in two ways. Firstly, Ca reacted vigorously in the presence of water and produced Ca(OH)₂ with a higher equilibrium pH in solution. Secondly, as mentioned earlier, in the dendritic microstructure, the Mg₂Ca phase was homogeneously distributed within the matrix and grain boundary. The higher Ca concentration near the grain boundary region might have provoked more active local electrochemical cells to activate the corrosion process with faster release of (OH)⁻ ions in the form of Mg(OH)₂, which in turn might have increased the pH of the solution. Compared with the others, Mg–5.0Ca exhibited the highest pH value, which was associated with a poor protection layer formed on its porous surface. It is concluded from hydrogen gas evolution and pH data that the degradation rate elevates with increase in Ca concentration. Our results also revealed that the degradation was massive in PBS solution compared with SBF or Hanks solution, as reported previously in Mg–Ca alloy systems. ¹⁰

Corroded surface studies

Morphological features, composition, and phase structure of Mg–0.5Ca and Mg–5.0Ca alloys immersed in PBS solution for 24 h are shown in Figure 4. The surface of Mg–0.5Ca seemed to be covered fully by a white corrosion film of insoluble Mg(OH)₂ which was evident by EDX (Figure 4(c)) and XRD studies of the corroded surface (Figure 5). Apart from white precipitates, cracks were also seen on the surface which might be due to layer dehydration in air. The degradation occurred due to the following electrochemical reactions in aqueous solution: ²⁴

OPEN IN VIEWER

Figure 4.

SEM images of corroded Mg–Ca alloy (a) 0.5 wt% Ca, (b) 5.0 wt% Ca after 24 h of immersion in PBS solution, along with their EDS profiles (c, d).

OPEN IN VIEWER

Figure 5.

XRD of reaction products after immersion in PBS solution at 37°C for 24 h.

Anodic reaction _ : Mg ( s ) = Mg 2 + ( aq . ) + 2 e −

(1)

Cathodic reaction _ : 2 H 2 O + 2 e - = 2 OH - ( aq . ) + H 2 ( g )

(2)

Overall reaction : _ Mg 2 + ( aq . ) + 2 OH - ( aq . ) = Mg ( OH ) 2 ( s ) ( Layer )

(3)

As a result of deposition of a protective layer of Mg(OH)₂-xH₂O, the corrosion rate was decreased notably afterwards, in agreement with the steady pH after long incubation time (Figure 3(b)). Also, the protective layer attracted PO^3-₄, CO^2-₃, and Cl^- ions into the solution towards the surface of the sample, which in turn might have accumulated more OH^- ions. On the contrary, the Mg–5Ca alloy showed a large number of deep corrosion pits (region A) along with cracks on its surface which signified higher dissolution rate of the alloy. EDX studies revealed the existence of O, P, Ca, and Mg elements in both Mg–0.5Ca and Mg–5.0Ca alloys (Figure 4(c, d)). The higher amount of oxygen further contributed to the higher degradation rate of Mg–5.0Ca alloys (Figure 4(d)). The highest corrosion rate might be attributed to the presence of an extensive amount of intermetallic Mg₂Ca phase in the vicinity of the primary Mg phase, which was clearly seen in Figure 1(d). The Mg₂Ca phase corroded more readily than Mg dendrite, leading to relatively faster degradation of the region around the primary Mg phase. The dissolution of this phase might have exposed the α-Mg phase, creating higher surface area with direct contact to PBS for enhanced reaction. Addition of up to 0.8 wt% Ca to the Mg matrix as a solid solution has been reported to lower the corrosion rate. ¹⁰ However, the presence of a non-homogeneous distribution of Ca with predominantly higher concentration along the grain boundary might have enhanced its degradation in PBS medium.

XRD patterns of Mg–0.5Ca and Mg–5.0Ca alloys after immersion in PBS for 24 h are shown in Figure 5. Results showed that Mg(OH)₂ (ICDD No: 01-082-2453) was the main corrosion product in both alloys. In addition, peak intensities of Mg(OH)₂ were much more pronounced in Mg–5.0Ca alloys compared with those of Mg–0.5Ca alloys. Such apparent difference might be due to the larger amount of reactive Mg₂Ca phase associated with Mg–5.0Ca alloys compared with those associated with Mg–0.5Ca alloys (Figure 1). XRD results supported the results of the EDX studies (Figure 4(c, d)).

Figure 6 is a schematic diagram showing the corrosion model at the alloy/solution interface after exposure in PBS solution at 37°C for 24 h. The electrochemical mechanism occurred as mentioned in Equations (1)–(3). As soon as the surface of Mg–Ca alloy was exposed to the PBS solution, the intermetallic Mg₂Ca phase served as a cathode and resulted in the hydrogen evolution (Equation (2)) due to its greater reactivity than the primary α-Mg phase. ¹³ The dissolution of Mg also occurred at the anode site (Figure 6(a)), as explained by Equation (1). The product layer of magnesium hydroxide (Mg(OH)₂) was formed on the Mg–Ca surface (Equation (3)) because of the relevant alkalization in the vicinity of the surface (Equation (2)). The increase in corrosion rate of Mg–5.0Ca, as depicted by SEM images (Figure 4), could thus be explained by the increment in cathode to anode area due to the presence of a larger amount of Mg₂Ca phase. Since the film on the surface of the Mg–Ca alloy was partially protective, as it was constituted of minute pores, the solution continuously penetrated via the film and reacted with the Mg–Ca alloy, resulting in an inward shifting of the reaction interphase. Simultaneously, the outward diffusion of dissolved Mg²⁺ ions occurred and produced Mg(OH)₂ at the surface. Thus, the Mg(OH)₂ layer growth was manifested by combined dissolution and precipitation mechanisms. The chloride ions present in solution disturbed this dissolution–precipitation process and transformed the insoluble Mg(OH)₂ into soluble MgCl₂ by combining with Mg²⁺ ions.^25–27 Certain areas of the Mg(OH)₂ film immediately dissolved resulting in pit formation, leaving the fresh alloy surface exposed to the solution. The same chemical reactions (Equations (1)–(3)) continued until the alloy was fully exhausted (Figure 6(b)). Hence, the severe non-uniform corrosion was revealed by the bulk Mg–Ca alloy until the Mg₂Ca phase was depleted.

OPEN IN VIEWER

Figure 6.

Schematic diagram showing the corrosion mechanism at alloy/solution interface (a) Early stage where galvanic corrosion initiates between Mg and intermetallic Mg₂Ca phase on exposure in PBS solution; (b) Severe corrosion with pit formation after immersion in PBS solution for 24 h.

Optical observation and micro-CT

After closing the defect site (femoral condyle) with suture, a gas pocket formed in the vicinity of the implant site in all rabbits. The size of the gas pocket increased rapidly in the initial 2 days but eventually stabilized within 1 week of implantation. Photographs of the surgical sites were taken prior to extraction of bone tissue containing the Mg–Ca alloy implants and are shown in Figure 7(A). A subcutaneous gas accumulation was seen after implantation of Mg alloys containing different contents of Ca into femoral condyle after 2 and 4 weeks. Tissue swelling was gradually reduced in all implants 2 weeks after the implantation and became significantly condensed after 4 weeks. However, a subcutaneous bulging or tissue swelling was observed to be smaller for the Mg–0.5Ca alloy than that of Mg–5.0Ca after implantation for 2 weeks (Figure 7(A,a), 7(A,c)) and 4 weeks (Figure 7(A,b), 7(A,d)). Upon extraction, the implant site was found to have accumulated high amounts of interstitial fluid. For in situ visualization of the implanted Mg–Ca alloys defect model, a transverse cut of the femoral condyle along the diameter of the implanted alloys was made. Figure 7(B) shows the optical images of implanted Mg–Ca alloys with different Ca contents after 2 and 4 weeks. In the Mg–0.5Ca alloy system, a gradual degradation was observed. A portion of the implant still remained after 2 weeks (Figure 7(B,a)) and 4 weeks (Figure 7(B,b)) of implantation. The degradation product (Mg(OH)₂) was also found to be accumulated on the surface of the implants. Between host bone and the implant, a sheath of cellular deposition appeared as tissue response was increased in thickness from 2 weeks to 4 weeks. However, the Mg–5.0Ca alloy was seemed to be almost totally degraded in all observation periods (Figure 7(B,c), Figure 7(B,d)).

OPEN IN VIEWER

Figure 7.

A. Photographs of subcutanious gas accumulation with the Mg–Ca alloy implanted femoral condyle after 2 (a,c) and (b,d) 4 weeks. B. Optical images for Mg–Ca alloy 0.5 wt% Ca (a,b) 5.0 wt% Ca (c,d) after 2 and 4 weeks.

Micro-CT images of varied Mg–Ca alloys implants after 2 and 4 weeks are shown in Figure 8(A). Both alloy types showed a decrease in defect size after 2 and 4 weeks. Minimal bone formation was observed in Mg–5.0Ca compared with that in Mg–0.5Ca after 2 weeks of implantation (Figure 8(A,a), Figure 8(A,c)). A single data slice taken from the micro-CT scans of the tissue samples showed that Mg–0.5Ca alloy implants resulted in visibly better bone formation compared with Mg–5.0Ca after 4 weeks (Figure 8(A,b), Figure 8(A,d)). Calculated BV/TV percent values as computed from micro-CT images are shown in Figure 8(B). After 2 and 4 weeks of implantation, BV/TV percent values for Mg–0.5Ca were found to be 1.57% and 34.13%, respectively. For Mg–5.0Ca, they were 0.79% and 4.88%, respectively. Although Mg–5.0Ca showed significant increase in bone formation after 4 weeks, Mg–0.5Ca was still extensively superior, having almost 10 times better bone formation at the same observation period. The higher bone formation by Mg–0.5Ca implantation might have resulted from its higher corrosion resistance, leading to significant reduction of degradation products and lowered inflammatory reaction. Figure 8(C) represents the overall degradation rate of Mg–0.5Ca and Mg–5.0Ca alloys after 2 week and 4 weeks of implantation. The percent degradation rate as obtained from micro-CT analysis for Mg–0.5Ca alloy was 63% and 49% after implantation of 2 and 4 weeks, respectively. However, due to the complete dissolution of Mg–5.0Ca alloys within 2 weeks, the degradation rate was found to be negligible (~1%) after 2 weeks and 4 weeks of implantation. The complete degradation of Mg–5.0Ca alloys may account for the severe corrosion that took place due to the existence of large quantities of intermetallic Mg₂Ca phase in the alloy system, as evident from the morphology (Figure 4(b, d)) and XRD (Figure 5) results of the corroded alloy.

OPEN IN VIEWER

Figure 8.

A. Micro-CT images of the Mg–Ca alloy implanted femoral condyle after 2 (a,c) and 4 (b,d) weeks, B. Percent bone volume (BV/TV), C. Degradation volume (percent) after implantation computed from micro-CT images.

Histological staining

Micrographs were taken from stained tissue sections to better visualize and analyze the corresponding tissue response. Results of histological analysis of varied Mg–Ca alloys implanted in femoral condyle for 2 weeks and 4 weeks after H&E staining are shown in Figure 9. No visible remnants of the alloy implants were found in the stained sections. Decalcification of the tissue samples in diluted nitric acid might have effectively dissolved all residual Mg and deposited Mg(OH)₂ near the interface. Low-magnification images of the stained sections from Mg–0.5Ca and Mg–5.0Ca implants are shown in Figure 9(a) and Figure 9(d). A buffer interaction zone between the implant and host bone was observed. The blank space seen in the central region of Mg–5.0Ca is the space occupied by the alloy implant (Figure 9(d)).

OPEN IN VIEWER

Figure 9.

Histological analysis of varied Mg–Ca alloy implanted femoral condyle stained by hematoxylin and eosin after 2 weeks and 4 weeks.

Higher magnification micrographs of both implant types after 2 weeks are shown in Figure 9(b) and Figure 9(e). Copious macrophage cells were observed in Mg–0.5Ca and Mg–5.0Ca alloy implants (Figure 9(b, e)). However, active osteoblast cell linings can still be seen along the implant interface forming new bone. Interestingly, large highly basic particles can be seen along the interface of the Mg–0.5Ca predominantly occupied by dense fibrous tissue. Improved bone formation was observed in both samples 4 weeks after implantation (Figure 9(c, f)). Macrophage activity was present along the implant interface and near the host bone after 4 weeks. The initial immune response to the defect and the aggressive implant surface due to the release of Mg(OH)₂ were subdued largely in a 4-week timespan with the 0.5 wt% Ca implant. The formation of these degradation products led to a heightened inflammatory response of the surrounding tissues. Prolonged inflammation was more pronounced in Mg–5.0Ca. Staining results showed that the Mg alloy with lower Ca content (Mg–0.5Ca) demonstrated better bone tissue regeneration compared with a similar sample with higher Ca (Mg–5.0Ca). In vivo degradation of this Mg alloy was also rather quick. Formation of the gas pocket observed at the beginning of the recovery period and the presence of accumulated interstitial fluid in the implant site is a direct response to the rapid build-up of corrosion byproducts. Although the tissue sections showed that the bone tissue along the margin of the defect was not considerably altered, bone formation did not substantially extend within the space occupied by the implants. Based on these results, bone formation was highly influenced by the stability of the alloy implant. The Mg–0.5Ca alloy was perceptibly more stable than the Mg–5.0Ca alloy, thus it was able to function as a support for bone formation.