In vivo bioactivity of a mineral based orthopaedic biocement.
IntroductionTo achieve a long lasting bone replacement, stable bone-implant interfaces are required. The majority of the cementation is made using PMMA. However the interfacial stability is low for PMMA due to high curing temperature and lack of bioactivity. To deal with these shortcomings many cements have been developed, e.g. glass polyalkenoate and various composite cements of calcium phosphate and polymers [1]. Biomaterials that do form a bond with tissue are called bioactive; the first material to show bioactivity, besides Apatite, was the bioglass [2]. Other materials with similar properties have also been developed, e.g. the Wollastonite ceramics [3]. Most bioactive materials cannot be shaped in vivo and have to be machined to desired shape. Calcium phosphate cements are bioactive and can be shaped intraoperatively. Their mechanical strength is however low and the material cannot be used in load bearing applications. The calcium phosphate cements are also resorbable. The present development of cements is divided into two major paths; calcium phosphate cements with increased properties (i.e. faster resorbtion rate or better mechanical properties) and composite materials to gain combined high mechanical strength and bioactivity (e.g. bioglass combined with polymers) [1]. A third solution is to explore other injectable bioactive minerals than the calcium phosphates, minerals from the so-called chemically bonded ceramics family [4]. Injectable bioactive minerals, which provide immediate stability, harden with limited exothermal reactions and are strong enough to allow for an early and active rehabilitation will allow function to be restored more rapidly [5-8].
In this study a novel biocement based on the calcium aluminate minerals Marokite and Grossite for primarily orthopedic applications was investigated in an animal study. The minerals were used as a powder, which upon mixing with water hardens through a reaction between the powder and the water. Due to the high amounts of water involved in the reaction, a high degree of mouldability is achievable, and a dense body of low residual porosity and high strength is possible to form [9]. When mixing with water the minerals dissolve and new minerals precipitate according to:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (1)
The precipitated minerals (hydrates) have a very fine grain size and a high surface energy with considerable amounts of bound water. The hydrates bond the material together and strength develops. The compressive strength reaches values close to 200 MPa and the flexural strength 50 MPa [10]. The biocement composition has been evaluated in in vitro biocompatibility testing according to the ISO 10993 standard prior to the animal tests. The histological response to the material has been reported elsewhere [11]. Therefore, the primary objective of this study was to investigate bone (cortical and cancellous) integration and generation of new bone in contact zones to the implanted material after a period of 6 weeks.
Materials and Methods
The biocement is composed of Grossite (CaO.2[Al.sub.2][O.sub.3]) and an isostructural form of Marokite, where [Fe.sub.2] [O.sub.3] is exchanged for [Al.sub.2] [O.sub.3], i.e. CaO x [Al.sub.2][O.sub.3]. The ratio of calcium--to aluminium oxide is approximately 5/11. For purposes of microstructure optimization, the mineral powder further contains calcium titanate (CaO x Ti[O.sub.2] as inert filler, which disperses well into the formed hydrates, and small amounts of calcium and silicon oxides. The microstructure is shown in Fig. 1.
The preparation of the mineral powder includes mixing by milling using silicon nitride balls in a polyethylene container with iso-propanol. The alcohol is evaporated and polymeric residues are removed in a furnace at 400[degrees]C. The powder mix is sterilized by electron beam radiation, and the aqueous solutions are steam sterilized. In order to prepare the samples, powder is mixed with two aqueous solutions. First 9.5 ml of deionized water containing small amounts of a plasticizer and a thickener is added to 30 g of the powder mix and a paste is prepared. Then 0.8 ml of 8.5 g/l LiCI solution is added for accelerated curing. After the accelerator solution is added the paste is mouldable for approximately 15 minutes at room temperature. The injectable paste was evaluated in an animal model involving 20 female albino adult (body weight about 2.5 kg) New Zealand White rabbits. The Local Ethics Committee, Malmo University, approved the experiments.
[FIGURE 1 OMITTED]
The paste was injected into pre-drilled holes, 5 mm in diameter and approximately 12 mm deep, in the tibia condyle of one or both rear legs. Autopsy took place after 6 weeks. Sections including both the bone and the implant were cut. The surgical procedure followed standard techniques. Polymethylmetacrylate (PMMA) based bone cement (CMW 1 from Johnson & Johnson) was used as reference material.
After necropsy and stabilisation in formaldehyde, the samples were imbedded in an acrylate polymer, cut and polished. The sample preparation procedure principally involves: Dehydration for one week in 70 %, 95 % and 98 % ethanol, respectively; infiltration for one week in glycolmethacrylate/alcohol mixes of increasing concentrations, ending with pure glycolmethacrylate, respectively; and finally polymerisation with UV--light. The specimens were split with a diamond saw blade and the cut surface polished. The microstructure of the implantlbone interface was studied with scanning electron microscopy (SEM), transmission electron microscopy (TEM, Jeol 2010 F) with a scanning module (STEM). SEM was used for overview studies of the interface, whereas TEM was used to retrieve information regarding the structure and elemental distribution in the contact zone. The compounds formed at the interface between the bone and the filling material were analyzed with energy dispersive spectrometry (EDS) mapping and line profiles in STEM-mode [12]. The TEM imaging was performed with a Gatan DualVision camera and STEM was performed with Gatan BF and ADF detector. The EDS-equipment was from Oxford Instrument with an INCA software. Cross-section TEM samples from the interface between bone and the bone biomaterial were produced using FIB [13]. To produce the TEM samples in this investigation the so-called "H-bar lift-out" technique was used enabling a very high site-specific accuracy 14]. The thickness of the TEM samples in this investigation was about 150 nm.
Results
The interface between bone and the mineral paste was continuous with no gaps, Fig. 2. It is notable that there is no calcium titanate closer than 10 Jlm from the bone. This indicates that there has been dissolution of Marokite and Grossite and precipitation of hydrates taking place during the formation of the contact zone. In Fig. 2a it is shown how new bone has grown towards the biomaterial and Fig. 2b is a close up from Fig. 2a. The reference material PMMA showed an irregular interface with gap between the bone and cement from a couple of microns up to several hundred microns.
A surface zone with a different composition than that of the bulk composition was detected at the biocement, Fig. 3. The thickness of the zone was about 200 [micro]m . The surface zone seemed to have a less dense structure and a higher porosity than the bulk material. In general there was more C and P and less Cl and Ca in the zone than in the bulk, see Figs. 3b and c. As can be seen in Fig. 3a there were distinct bright areas within the surface zone. These areas contained almost exclusively Ca, O and C, see Fig. 3d.
[FIGURE 2 OMITTED]
More structural information was obtained by using TEM, Fig. 4. Towards bone, three different zones were detected, the original bone structure (left in Fig. 4a), a more dense bone structure (darker areas towards the cement marked as a) and a different less crystalline structure corresponding to the cement (bright areas marked as B). In the image some sample preparation cracks and a pore is also visible as very bright. In electron diffraction, the darker bone structure had larger crystal lites (i.e. it scattered more in bright field imaging) than the original bone (rings). The diffraction patterns from the dark bone structure contained both elongated dots and rings whereas the bright cement areas showed very weak rings, Fig. 4c and d. By using dark field imaging of crystalline Apatite, its location in the biocement could be detected, Fig. 4b.
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
The transition from bone to biocement was continuous and the actual interface was difficult to resolve using TEM, Fig. 5, where the bone to the left in the image was less dense and contained slightly more P than the bone to the right. Areas of Gibbsite seemed to be completely surrounded by the "denser" bone. All Ca could be connected to P, indicating that no Katoite had been formed, see Fig. 5c-e. The interface between Gibbsite and bone tissue was free from gaps.
[FIGURE 5 OMITTED]
Discussion
The microstructure of the cement is different at the interface to bone compared to that in the bulk. At the interface the minerals Apatite, Calcite and Gibbsite were found, whereas in the bulk Gibbsite and Katoite were present. It is also important to notice the morphology and separation of the minerals at the interface. Two different interfaces were found, the interface towards bone marrow and towards bone tissue.
Towards bone tissue, the Apatite had a "bone structure" and was separated from the Gibbsite. The bone formed in close vicinity to or in the biocement was denser and more crystalline than the bone formed some microns away from the interface. Amorphous Gibbsite "islands" were surrounded by dense bone (as demonstrated from the diffuse rings in the diffraction patterns). Although different from the morphology of the bone formed at a distance from the interface this bone tissue seems to have a similar fibrous structure, indicating that cellular activity has participated in its formation. How and what type of cells that interact with the material to form the "bone structure" in the biocement is unknown. For bioglasses, the understanding of the cell response to the material is profound [15], and thus much of that knowledge could probably be transferred and adopted to the bioactive minerals.
At the interface towards bone marrow, a surface zone divided in Gibbsite and Calcite areas was found. Small amounts of phosphorus indicated that also Apatite could be present in the zone. Compared to the interface towards bone tissue different phases were more distinct in location, with Ca bound as Calcite in certain areas and Al bound as Gibbsite.
From a chemical point of view the difference in mineral composition at the interface compared to the bulk can be understood from considering the minerals solubility products. In
the bulk the supply of phosphate ions is limited and the hydration reaction (1) occurs. The pH of the hydration reaction for Marokite is about 11. At the interface where the supply of phosphate ions is considerable calcium phosphate minerals can also be formed. Given the high pH, Apatite is the most stable calcium phosphate mineral [16]. From the solubility products for Katoite pKs = 22.3 [17], Gibbsite pKs = 32.3 [18] and Apatite pKs = 57.8 [16] and the number of ions involved in the formation of each mineral it is clear that Apatite will be formed together with Katoite. The solubility of Gibbsite is in the order of 10 nmol/l compared to 300 nmol/l for Apatite. Gibbsite is stable down to a pH of 3 [18]. Due to the high pH, leakage of [Al.sup.3+] is chemically not feasible, instead the low soluble Gibbsite mineral is formed via aluminate Al[(OH).sub.4]-ions.
Thus it can be hypothesized that what is observed is in fact Katoite dissolving at the surface in a Ca and phosphate ion acid-base mediated reaction occurring at near neutral pH:
[Ca.sub.3]. [(Al[(OH).sub.4]).sub.2]. [(OH).sub.4] + 2 [Ca.sup.2+] + HP[O.sub.4.sup.2-] + 2 [H.sub.2] P[O.sub.4.sup.- ] = > [Ca.sub.3]. [(PO4).sub.3]. (OH) + 2 Al[(OH).sub.3] + 5 [H.sub.2]O (2)
In this reaction Apatite and Gibbsite are produced. This process must take place in several steps each involving the formation of one water molecule. In tissues containing physiological fluids rich in pH buffering ions such as phosphate and hydrogencarbonate, it can be envisioned that these ions increase the rate of sub-micron AH3 precipitation out of this colloidal gel. As this reaction proceeds more phosphate and hydrogencarbonate ions are consumed and converted into PO43--and CO32-. Since calcium ions are present both in the extracellular fluid as well as in the solution surrounding the cement paste and pH is above 8, the prerequisites for forming calcium carbonate (Calcite) and Apatite are met. Since the solubility products for Calcite and Apatite are much lower than that of Katoite at pH above 8, Calcite and Apatite are precipitated to a larger extent than Katoite at the implant to tissue surface zone.
This paper focuses on reactions that have occurred within the first 6 weeks. Although not presented here, studies of the material after 6 months in-vivo shows the material to be stable. Thus the interface seem to be formed early in the process, directly after insertion, and then essentially to be unchanged over time. In vitro tests of long-term compressive strength have proven the material to increase its strength during the first month and then stay constant.
Conclusion
An injectable biocement based on Marokite was implanted into drilled cavities in the tibia of rabbits and evaluated regarding the interface between bone marrow, new-formed bone tissue and the biocement. Towards bone marrow the interface was heterogeneous with distinct areas with Gibbsite and Calcite. Electron microscopy and elemental analysis reveal a continuous Apatite containing region from the original implant to new bone with absence of gaps. Since the material forms a bond to living tissue it can be considered bioactive.
References
[1.] Kenny S. M., Buggy M., Bone cement and fillers: A review, Journal of Materials Science: Materials in Medicine, 14 (2003) 923-938
[2.] Hench L. L., Biomaterials: a forecast for the future. Biomaterials 1998;19:1419-1423. 3. Liu X., Ding C., Wang Z., Apatite formed on the surface of plasma-sprayed
[3.] wollastonite coating immersed in simulated body fluid. Biomaterials 2001;22:2007-2012.
[4.] Ravaglioli A., Krajewski A., Bioceramics, Chapman&Hall, 1992
[5.] Roemhildt M. L., McGee T. D., Wagner S. D., Novel calcium phosphate composite bone cement: strength and bonding properties, Journal of Materials Science: Materials in Medicine, 14 (2003) 137-141.
[6.] Engqvist H., Schulz-Walz J-E., Loof J., Botton G. A., Mayer D., Phaneuf M. W., Ahnfelt N-O., Hermansson L.: Chemical and biological integration of a mouldable bioactive ceramic material capable of forming Apatite in vivo in teeth, Biomaterials, 25 (2004) 2781-2787.
[7.] Seung-Han 0., Se-Young C., Yong-Keun L., Kyoung N. K., Seong-Ho C., Effects of litium fluoride and maleic acid on bioactivity of calcium aluminate cement: Formation of hydroxyapatite in simulated body fluid, J Biomed Mater Res, 67A-104-111 (2003).
[8.] Abdullah D., Pitt Ford T. R., Papaioannou S., Nicholson J., McDonald F., An evaluation of accelerated Portland cement as a restorative material, Biomaterials, 23 (2002) 4001-4010.
[9.] Scrivener K., Capmas A., Calcium Aluminate Cements, Lea's Chemistry of Cement and Concrete. Edited by P .C. Hewlett, Arnold 1998.
[10.] Axen N., Persson T., Bjorklund K., Engqvist H., Hermansson L., An Injectable Bone Void Filler Cement Based on Ca-Aluminate, Key Engineering Materials Vols. 254-256(2004) pp.265-268.
[11.] Axen N., Bjursten L. M., Engqvist H., Ahnfelt N-O, Hermansson L., Zone formation at the interface between Ca-aluminate cement and bone tissue environment, Proceedings of9'th Ceramics cells and tissue, Faenza Italy 2004
[12.] Williams D. B., Carter C. B., Transmission electron microscopy, Plenum Press 1996
[13.] Phaneuf M. W., Applications of focused ion beam microscopy to material science specimens. Micron 1999; 30: 277-288.
[14.] Patterson R. J., Mayer D., Weaver L., Phaneuf M. W., "H-Bar Lift-Out" and "Plan-View Lift-Out": Robust, Re-thinnable FIB-TEM preparation for ex-situ cross-sectional and plan-view FIB specimen preparation. Submitted to Microscopy and Microanalysis.
[15.] Hench L. L., Stimulation of bone repair by gene activating glasses, Key Engineering materials, Vols. 254256 (2004) 3-6.
[16.] Narasaraju T. S. B., Phebe D. E., Some physico-chemical aspects of hydroxylapatite, Journal of Materials Science 1996;31:1-21.
[17.] Menetrier-Sorrentino D., George C. M., Sorrentino F. P., The setting and hardening characteristics of Calcium Aluminate cements: Studies of the systems CI2A7-C3A and CI2A7-CA, Proceedings of International Congress of the Chemistry of Cement, Rio de Janeiro (1986).
[18.] Scriverner K.L., Cabiron J-L., Letoumeux R., High-performance concretes from calcium aluminate cements, Cement and concrete research 29 (1999) 1215-1223.
Hakan Engqvist (1) *, Martin Couillard (2), Gianluigi A. Botton (2), Mike W. Phaneuf (3), Niklas Axen (1), Nils-Otto. Ahnfelt (1), L. Hemlansson (1)
(1.) Doxa AB, Axel Johanssons gata 4-6, SE-754 51 Uppsala, Sweden
(2.) Brockhouse Institute for Materials Research, McMaster University, Hamilton, ON, L8S 4Ml, Canada
(3.) Fibics Inc, 568 Booth Street, Suite 224, Ottawa, Ontario, KIA OGl, Canada
* Corresponding author
 Printer friendly
 Cite/link
 Email
 Feedback
| |
| Author: | Engqvist, Hakan; Couillard, Martin; Botton, Gianluigi A.; Phaneuf, Mike W.; Axen, Niklas; Ahnfelt, N |
|---|---|
| Publication: | Trends in Biomaterials and Artificial Organs |
| Date: | Jul 1, 2005 |
| Words: | 2817 |
| Previous Article: | Chronology of total hip joint replacement and materials development. |
| Next Article: | Thermodynamic study of bone biocomposites. |
| Topics: | |