Calcium phosphate materials containing alumina: Raman spectroscopical, histological, and ultrastructural study A. BertoluzzaTt R. Simoni? and A. Tinti*
Centro Studi Interfacoltli sulla Spettroscopia Raman, *Sezione di Chimica e Propedeutica Biochimica, Dipartimento di Biochimica, via Selmi 2, 40126 Bologna, #Dipartimento di Chimica “G. Ciamician: via Selmi 2, 40126 Bologna M. Morocutti, V. Ottani, and A. Ruggeri istituto di Anatomia Umana Normale, via Irnerio 48, 40226 Bologna, Italy Variable alumina quantities were added to two types of calcium phosphate materials -hydroxyapatite ceramics with Ca/P = 1.67 and calcium metaphosphate glass with Ca/P = 1-in order to increase their mechanical properties. Raman spectroscopy shows that alumina interacts with the phosphate group of these materials, while thermomechanical analysis shows that their elastic modulus has a value similar to that of bone. Histological sections demonstrate that the surface in close contact with ceramic materials
shows a good integration between bone and biomaterial. All ceramic specimens are penetrated by well-stained, presumably glycoprotein, matrix, that consistently forms a thin network in close contact with the implant. After 6 weeks bone growing on both ceramic and glass shows signs of maturation with a lamellar structure and an apparently normal mineralization. In the case of the glass, inside this newly formed bone, were often observed two layers, the internal one showing a well defined lamellar structure.
INTRODUCTION
Over the last several years efforts have been made to develop bone implant materials composed of calcium phosphate compounds because their chemical composition is closely allied to the mineral component of bone. Implant materials composed of calcium phosphate display excellent biocompatibility under many conditions and different forms of these implant materials have been studied in animals and The most widely clinically applied calcium phosphate ceramics are composed of either hydroxyapatite Ca,(PO,),(OH) (HA) or P-tricalciumphosphate Ca3(P04)2 (p-TCP) both with more or less successful results. Among materials containing calcium phosphate, but also with other components in certain quantity, are certain glasses and glass-ceramics.3,5’o ‘To whom correspondence should be addressed. Journal of Biomedical Materials Research, Vol. 25, 23-38 (1991) 0 1991 John Wiley & Sons, Inc. CCC 0021-9304/91/010023-16$04.00
24
BERTOLUZZA ET AL.
With regard to HA, depending on the function required by the implant site, there are several different forms of dense and porous hydroxyapatite available for bone implantation, each with characteristics that recently have been extensively st~died’~~~~’-” and which depend on the way the ceramic was prepared. A particular property of some calcium phosphate materials is their biodegradability. The factors which determine bioresorbability of ceramics are still controversial. However, it is generally thought that chemical composition, crystallographic structure, density, and porosity (in particular microporosity) can be identified as main parameter^.'",'^-'^,^^,^' Because of their apparent ability to bond directly to bone, these materials are called bioactive materials (i.e., hydroxyapatite, P-tricalciumphosphate, bioglasses, and glass-ceramics). Although not yet fully understood, the existence of this bond, and to some extent its appearance and composition, has been documented by numerous investigators using a variety of materials and animal models?,9.11,22-24 Despite their useful properties like biocompatibility and bone bonding, the main restriction of some calcium phosphate implant materials is their poor mechanical properties. Many investigations have been carried out in order to induce reinforcement of mechanical properties; some of these adding various oxides, like alumina, obtaining results not always e f f e ~ t i v e . ’ ~ ” * ~ ~ ~ ~ ~ The aim of the present study is to investigate the effect of the addition of alumina to hydroxyapatite and calcium metaphosphate glass and to determine bone tissue interaction when implanted in the tibiae of rats. MATERIALS A N D METHODS
Calcium hydroxyapatite (HA) powder with the composition CaS(PO&OH, supplied by Merck Co., was used for the preparation of ceramic materials. X-ray analysis on this product is described in the l i t e r a t ~ r e ’ ~ ”and ~ ~ shows ’~,~~ peaks of the hydroxylapatite structure. The powder was compressed uniaxially at 300 MN/m2 and sintered at 1200°C for 1 h. Then the ceramics were ground and newly sintered twice to improve the ductility of the material and also to obtain a greater homogeneity of the samples. Samples containing alumina (Erba product) were prepared in the same way. The porosity of the ceramics was calculated from the weight and volume of sintered samples. Total porosity ranged between 35% (pure HA) and 45% (ceramics containing alumina). Calcium metaphosphate glass, Ca(P03)2,was obtained by melting the crystalline metaphosphate P-Ca(P03)2at 1300°C for 1 h in a platinum crucible and then quickly cooling it. /3-Ca(P03)2was prepared by dehydration of monocalcium phosphate monohydrate (Fluka product), according to the 1iteratu1-e.~’ Glasses containing alumina were made by means of the same thermal process starting from monocalcium phosphate monohydrate (Ca(H2P04)2 . HzO) and alumina (A1203)so as to obtain the required ratio.
CALCIUM PHOSPHATE MATERIALS CONTAINING ALUMINA
25
The composition of the samples is listed in Table I. Ceramic (0.8 X 0.5 x 1.5 mm) and glass (dia. 0.5-0.8; 1 mm length) samples were implanted respectively into the right and the left tibia of the same Wistar rat (200 g average weight). Three animals for every two specimens were operated reaching a total of 15. Rats were anesthetized by intraperitoneal injections of sodium pentobarbital. The incision sites were shaved and cleaned with ethanol. Holes were drilled employing a diamond burr (dia. 0.8 mm) continuously cooled with physiological saline. After careful removal of bone debris, implants were inserted (press-fitted) and the wound closed by suture. Control holes were prepared the same way, but no implant was inserted. Animals were sacrificed 3, 4, and 6 weeks after operation. The segments of the tibiae were excised and immediately immersed in a 10% buffered formalin solution for histological processing. Because of the low solubility of the glasses into the decalcifying solution employed, it was necessary to remove them from the implant sites, and this was done after fixation to ensure a better sectioning. Hydroxyapatite implants were dissolved by demineralization into a solution of sodium citrate and formic acid, dehydrated in alcools and embedded in paraffin. Longitudinal and transversal histological sections were stained with hematoxylin-eosin or Azan Mallory stains. Raman spectra of glass and ceramics before and after implant into rat tibiae were recorded using a Jasco R 300 spectrometer and 600 mW of 488.0 nm radiation from a Spectra Physics argon ion laser. Dynamical load thermomechanical measurements (TMA) were carried out by means of a Mettler TA 3000 system equipped with a TMA 40 accessory at 37°C. Scanning electron microscopy (SEM) was carried out using a Phillips 501 electron microscope: one sample of every kind of material was mounted using colloidal silver and coated in argon atmosphere with a gold and palladium layer. TABLE I Ca/P and P/Al Ratio and Corresponding A1203Weight Percent of Vitreous and Ceramic Materials Studied Samples
Ca/P 1 1 1 1 1 1.67 1.67 1.67 1.67 1.67
u = vitreous, c = ceramic materials. "Sample 5 is not well vitrified.
PfA1
AlzOa weight% (w/w)
2.51 4.90 6.85 9.33 1.49 2.95 4.18 5.73
BERTOLUZZA ET AL.
26 RESULTS
Vibrational Raman spectroscopy Table I1 shows spectroscopical modifications of hydroxyapatite-alumina ceramics with various P/A1 ratio. As the alumina content in the ceramic increases, new bands at about 520,555,630, and 1090 cm-' appear in the spectra, and their intensity increases as alumina content does. These new bands are not present either in the alumina or hydroxyapatite Raman spectra. Table I11 shows spectroscopical modifications of calcium aluminophosphate glasses with various P/A1 ratio. As the alumina content increases, an increase in the frequency of the strong Raman band at about 690 cm-' in the calcium metaphosphate glass is observed. This band rises to 711 cm-' for the glass with 5/1 P/A1 ratio. On the contrary, there is a frequency decrease, less evident, of the strongest band at 1178 cm-' for calcium metaphosphate glass. This band is centered at about 1167 cm-l for the glass with the higher alumina content. A band practically absent in the spectrum of calcium metaphosphate glass appears in the spectra of calcium aluminophosphate glasses at about 1025 cm-' and augments in intensity increasing the alumina content of the glass. Thermomechanical analysis (TMA) Figure 1 shows TMA measurements performed on some glasses and ceramics and on rat tibiae (measured exactly in the same site in which the implant was performed). It is evident from the figure that glass samples have a TABLE I1 Raman Bands of Ceramic Materials HA 380 (39) 417 (100)
580 (10) 645 (30) 750 (36)
427 (14) 450 (sh) 578 (10) 590 (10) 604 (6) 947 (sh) 960 (100) 1025 (5) 1045 (8) 1075 (4) -
P/Al
=
2011
408 (sh) 430 (14) 450 (sh) 550 (2) 580 (10) 590 (10) 608 (7) 630 (sh) 947 (sh) 960 (100) 1025 (5) 1045 (8) 1075 (4) -
P/A1 = 10/1
410 (6) 430 (14) 450 (sh) 525 (sh) 552 (2) 580 (10) 592 (11) 610 (7) 630 (sh) -
947 (sh) 960 (100) 1025 (5) 1045 (7) 1075 (5) -
410 (6) 430 (14) 450 (sh) 525 (sh) 555 (2) 580 (10) 592 (11) 610 (7) 630 (4) 947 (sh) 960 (100) 1025 (sh) 1045 (7) 1075 (5) 1090 (sh)
sh = shoulder; numbers in parentheses are the intensities of the bands.
410 (11) 430 (14) 450 (sh) 520 (sh) 555 (2) 580 (10) 592 (11) 610 (8) 630 (4) 947 (sh) 960 (100) 1025 (sh) 1045 (7) 1075 (5) 1090 (sh)
CALCIUM PHOSPHATE MATERIALS CONTAINING ALUMINA
27
TABLE 111 Raman Bands of Glass Materials Al2O3
Ca(PO&
P/AI = 2011
P/AI = 1011
P/AI = 711
P/Al = 511
-
327 (12)
335 (16)
340 (19)
345 (22)
345 (25)
380 (39)
-
417 (100) 580 (10) -
395 (10)
-
-
-
-
525 (5)
-
600 (2) -
-
-
397 (sh)
395 (13)
-
525 (6)
525 (6)
-
-
605 (3)
610 (3)
-
612 (5)
612 (7) 711 (75)
-
-
698 (75)
705 (75)
795 (sh) 935 (1) 1023 (3) 1083 (6) 1178 (100) 1277 (15) 1342 (sh)
795 (sh) 935 (sh) 1025 (8) 1085 (sh) 1176 (100) 1266 (19) 1330 (sh)
797 (sh) 935 (sh) 1025 (14) 1100 (sh) 1171 (100) 1265 (20) 1330 (sh)
797 (sh) 935 (sh) 1025 (18) 1105 (sh) 1170 (100) 1265 (20) 1327 (sh)
-
-
-
-
525 (8)
-
698 (69)
-
-
397 (sh)
525 (7)
693 (51)
750 (36) -
395 (14)
-
645 (30)
-
-
-
6POz 6POP
-
-
V,POP 798 (sh) 935 ishi 1025 (28)
j va,pop
sh = shoulder; numbers in parentheses are the intensities of the bands.
I
I
1
0
0.1
0.2 PROeE
0
DISPLACEMENT
0.1
(*%)
Figure 1. Dynamic load thermomechanical analysis, corrected for the blank, of: (a) A calcium aluminophosphate glass with P/A1 = 10/1 (sample 3), A ceramic with P/Al = 10/1 (sample C), 0 rat tibiae bone; (b) A calcium aluminophosphate glasses with P/Al = 7/1 (sample 4), A P/A1 = 10/1 (sample 3) and calcium phosphate glass (sample 1).
BERTOLUZZA ET AL.
28
relatively higher compressive strength than ceramic samples. Moreover, TMA measurements performed on glass samples (1,3, and 4)and on ceramics (A, C, and E) reveal some differences in elastic behavior as a function of the composition (percent of alumina) of either glass or ceramic materials. In particular, the addition of alumina increases the compressive strength. The Young modulus calculated from these measurements is of the order of 1.5.10' Pa for ceramics, 2.5.10* Pa for glasses and 6.107 Pa for rat bone.
Scanning electron microscopy Before implant the microstructure of HA blocks consists of regular and rounded particles connected by grain boundaries resulting from the sintering procedure. The microporosity measured by weight and volume of sintered samples ranges from 35% for pure HA to 45% for samples containing the greater amount of alumina (Fig. 2). Glass blocks show an almost fully smooth surface and an amorphous microstructure. Incidental fractures show a sharp-edged shape typical of glass.
Implantation
Ceramic Histological sections do not show any inflammatory reactions from the first time period on three weeks studied. On decalcified sections HA speci-
Figure 2. SEM showing the HA-A1203material (sample C) porosity. Original magnification X4750.
CALCIUM PHOSPHATE MATERIALS CONTAINING ALUMINA
29
mens show a "vacuolized" appearance presumably caused by the material grains dissolved with decalcifying solution. Specimens with alumina addition have a more microgranular pattern and a higher density with respect to pure HA (Fig. 3). Four weeks after implant both HA and HA with A1,03 specimens are still completely surrounded by primary bone laid also down on medullary surface. The narrow layer of primary bone with woven fibers still shows a clear separation from the preexisting lamellar bone (Fig. 4). Some concentric lamellae are also visible. The surface in close contact with the implant shows a good integration between bone and biomaterial. All the specimens are penetrated by well-stained, presumably glycoprotein, matrix that all along forms a thin network in close contact with the implant. At 6 weeks we can still find primary bone but its appearance is becoming more lamellar and more integrated with preexisting bone mostly in subendosteal and periosteal areas. Specimens with higher content of A1203 show an increased thickness of the stained matrix at the surface of the implant. In some cases inside larger
Figure 3. Longitudinal histological section of HA specimen after 6 weeks showing a vacuolized aspect. Implant is completely surrounded by primary bone. Azan-Mallory staining. I = implant. Original magnification ~ 3 0 .
30
BERTOLUZZA ET AL.
Figure 4. Longitudinal histological section of HA-A1203 material (sample E) showing a microgranular pattern. The specimen is surrounded by new bone also on medullary surface. 4 weeks after implant. AzanMallory staining. I = implant. Original magnification x8.
lacunae at the implant surface some thin connective branches are observable (Fig. 5).
Glass No significant differences were seen in comparison with the interactions of different compositional glasses with the surrounding tissue. Four weeks after the implant the specimens were already surrounded by newly formed bone apparently without any connective tissue at their interface. Only in a few samples a thin layer of stainable material, occasionally containing cell debris, was observable. It is also apparent that tissue adjacent to where the implant was removed was predominantly osteoid and was also surrounding the medullary canal. The boundary between old and newly formed bone was always clearly visible because of differences in stainability and lamellar orientation (Fig. 6). After 6 weeks bone growing on both glass with and without A1203addition showed signs of maturation which was represented by a prevalent substitution of primary bone with lamellar structure. Inside this newly formed bone it was frequently possible to notice the presence of two layers. On the one hand, bone directly bordering the external implant surface had more matured attaining a concentrical lamellar pattern. On the other hand the outer
CALCIUM PHOSPHATE MATERIALS CONTAINING ALUMINA
31
Figure 5. Longitudinal histological section of HA-A1203material (sample D) 6 weeks after operation showing connective branches at the implant surface. Azan-Mallory staining. I = implant. Original magnification ~ 1 8 5 .
layer of newly formed bone, not yet completely lamellar, still showed a boundary with the old bone, but a better integration was visible (Fig. 7). In control animals in almost all of the cases newly formed bone attains an aspect very similar to that of a normal secondary bone within 6 weeks.
DISCUSSION
TMA measurements show that the elastic modulus of our glass and ceramic materials is quite similar to that of rat tibiae. In both materials the increase of the Young modulus for addition of alumina suggests the possibility of changing the elastic characteristics of the material as a function of the requirements of different implant sites.
BERTOLUZZA ET AL.
32
Figure 6. Longitudinal histological section of glass (sample 2) 4 weeks after operation showing the clear separation between old and newly formed bone. (-) Azan-Mallory staining. Original magnification X50.
Ceramic
Concerning hydroxyapatite-alumina ceramics, Table I1 evidences that the band at 410 cm-' for ceramics with higher alumina content can derive from the more intense band at 417 cm-' in the alumina spectrum.31In the same way the new band at 630 cm-' can be related to the band at 645 cm-' in the alumina Raman spectrum. The new bands at 520 and 555 cm-' can be attributed to modifications in the structure of phosphate groups of HA induced by the interaction with A1203,as already reported by other authors for v4 stretching modes of phosphate groups in substituted apatite^.^^,^^ Moreover, the two bands attributed to phosphate groups at 578 and 590 cm-I only partially change their frequency owing to this interaction. A new weak band at about 1090 cm-', not present in the spectrum of alumina, is evident in ceramics containing more alumina. This can be related to the 1075cm-' band of hydroxyapatite. These spectral modifications show that the sintering process of ceramics has led to the synthesis of materials where the Raman bands of AlzO, and of the phosphate group are lightly modified. These changes can be related to a chemical interaction that occurred during the sintering process. Density measurements show that HA with alumina has a greater microporosity than normal HA. Moreover, SEM micrographs of the surface of HA with alumina samples show an apparent increase of microporosity, with a greater number of small pores.
CALCIUM PHOSPHATE MATERIALS CONTAINING ALUMINA
33
Figure 7. Transversal histological section of glass (sample 4) showing the presence of two layers: the one closer to the implant has a concentrical ) staining. Original magnification X 52. lamellar pattern. (4Azan-Mallory
The presence of A13+ in the sintered material leads to a delay in the grain growthz5and to a decrease in the sizes of the pores. From decalcified histological sections, the material of samples B, C, D, and E appears more microgranular, probably owing to alumina residual, after the material has been pervaded and stabilized by the organic component of the matrix. All the HA-implanted materials show a good tissue integration as demonstrated by the absence of any inflammatory or toxic reactions of the tissue in the vicinity of implants. Histological sections in contact with the implant surface typically show bone growing in direct contact with the implant material without any intervening fibrous tissue capsule. HA materials completely surrounded on all sides allow the formation of a good attachment between implant and bone. During histological processing HA is firmly adherent to the bone so that an eventual breakage preferably takes place between new and old bone rather than between new bone and implant. Raman spectra of ceramic materials after implant show only some stretching bands which can be attributed to organic matter, probably caused by organic phase of bone. The mineral phase of bone, nevertheless, cannot be observed because its bands are coincident with those of HA. The osteointegration and the progressive maturation of the bone surrounding the implant material is probably better for samples containing alu-
34
BERTOLUZZA ET AL.
mina than for pure HA. Samples used in our study are well pervaded by a substance which may be glycoprotein containing stainable material which on the surface becomes fairly thick and is almost organized into some superficial lacunae on small connective branches. This behavior contrasts with the hypothesis that addition of this oxide to Ca/P biomaterials normally has a negative effect on their bio~ompatibility?~,~~ The discrepancies in the results obtained are probably related to a difference in the percent of alumina employed or otherwise give rise to some doubts about the inhibitory effects of alumina.35 Glass
As shown in Table 111, addition of basic oxides to metaphosphate glasses produces a demolition of the polymeric structure and the formation of oligomers characterized by an intensity increase of the stretching vibration of PO:- terminal groups at about 1020-1040 cm-' and a decrease of 1170- and 690-cm-' bands of the polymeric structure.36 The spectroscopical trend for calcium metaphosphate glasses containing A1203is similar to that of oligophosphates, but the new band attributable to PO:- groups terminal of the chain is much less intense and localized at a lower wavenumber (1025 cm-I). On the contrary, the lowering of the 690 cm-' band, relative to POP groups, is of the same order as that observed for calcium oligophosphate glasses. This behavior can be interpreted considering that the presence in the chains of alumina breaks the vibrational coupling of POP groups, but the lower presence of PO?- groups suggests a partial covalent interaction A1-0 with respect to the prevailingly ionic interaction of calcium.37 Histological observations on rat tibiae where vitreous materials have been implanted show, in any case, a good response of the host tissue, and indicate an excellent tissue integration. calcium With regard to this item, as already reported by other metaphosphate gives good results, and also in agreement with Zhu et a1.6our glasses containing alumina are well tolerated by the tissue. Seemingly contradictory results were obtained by Daculsi and his when they showed that alumina added to Ca/P materials reduces their biocompatibility. Indeed their experimental conditions differ from ours, as their material was implanted in connective tissue rather than in bone. Moreover their ultrastructural observations are restricted to cell reactivity while our data include the environment of the host tissue in its globality. Our data also show the absence of signs of any inflammatory reaction, the absence of a fibrous capsule and, indeed, a generalized formation of new bone which extends on all sides of the implant, including the medullary one. Already 4 weeks after implant, at the interface with the glass, a bone lamellar structure is observable and this demonstrates an advanced stage of bone remodelling. One may suggest that the glassy material, possibly because of its elastic modulus (which is similar to that of bone), is effective in the sec-
CALCIUM PHOSPHATE MATERIALS CONTAINING ALUMINA
35
ondary bone formation. This behavior is also observable in samples containing 10% in weight of alumina. An addition of this kind of oxide would have an inhibitory effect on mineralization. In fact, these oxides are used to stabilize and delay the release of ions from vitreous material^?^*^^^^^,^^,^^,^^ The fact that alumina in our samples appears to have no influence seems to derive from the weak effect of this oxide or from the characteristic structure of the glass itself. Moreover, the absence of any visible acute inflammatory reaction should be evidence in favor of a surface controlled reaction of the vitreous material. The poor solubility of a vitreous material does not seem to be a sure parameter for the absence of bone since at the surface of the scarcely soluble material it is possible that after a long time an adsorption of proteins capable of promoting a deposition of a inorganic Ca/P layer could take place. In fact, the Raman spectrum of one of our glasses after implant revealed the presence of a weak band centered at about 960 cmP, which can be related to an apatitic phase on the surface of the glass. For other samples this band was not evident, probably because of the difficulties in obtaining the surface Raman spectra of small transparent materials. Maybe, the Ca/P rich layer is not by itself a sufficient reason for bone bonding under functionally loaded conditions. Possibly implant surface porosity or microstructure is more important than the effect of ions leached from the implant s ~ r f a c e . ~ ~ , ~ ’ Limited to 6 weeks, our samples do not show a great adhesion to the surrounding tissue and this behavior confirms the hypothesis that the presence of alumina in glass is detrimental to bone b ~ n d i n g . ~ ~However, , ~ ~ , ~ * , consider~’ ing the limited time of our experiment we cannot exclude that bone bonding takes place after a longer time, i.e., 12 weeks, as estimated elsewhere for nearly insoluble glass with addiction of a l ~ m i n a . ~ ~ , ~ ~ In conclusion, it has been observed that mechanical and elastic properties of calcium phosphate materials (i.e., ceramics and glasses) increase as the alumina content of the material increases. Regarding the materials with a hydroxyapatite structure there is a sintering process of alumina-phosphate evidenced by vibrational spectroscopy, and by Scanning Electron Microscopy. On the contrary, in calcium metaphosphate, alumina induces a breakage of the polyphosphate and formation of oligomers with shorter chains, where PO;- terminal groups are bonded with a partial covalent character to aluminium. As to the biocompatibility of calcium phosphate materials containing alumina, it has been observed that small quantities of alumina are well tolerated. Both implant materials are surrounded by newly formed bone that become progressively mature showing a good osteointegration in the case of ceramic material, and a formation of two bone layers in that of the glass. References 1. M. Jarcho, ”Calcium phosphate ceramics as hard tissue prosthetics,” Clin. Orthop. Rel. Res., 157, 259-278 (1981). 2. T. Han, F. A. Carranza, and E. B. Kenney, ”Calcium phosphate ceramics in dentistry: a review of the literature,” J. West. SOC.Periodont. Periodont. Ah., 32, 88-108 (1984).
BERTOLUZZA ET AL.
36 3. 4.
5. 6. 7.
8.
9.
10. 11. 12. 13. 14.
15.
16.
17. 18. 19.
20.
F. Pernot, R. Rogier, J. Zarzycki, F. Bonnel, and P. Baldet, “Les biomateriaux vitreux et vitroceramiques en chirurgie orthopedique et OSseuse,” Bull. SOC. Ckim. Fr., 4, 519-522 (1985). K. de Groot, ”Degradable ceramics,” in Biocompatibility of Clinical Implant Materials, Vol. I, D. F. Williams, (ed.), CRC Press, Boca Raton, 1981, pp. 199-222. L. L. Hench and A. E. Clark, ’Adhesion to bone,” in Biocompatibility of Orthopaedic Implants, Vol. 11, D. F. Williams (ed.), CRC Press, Boca Raton, 1982, pp. 129-170. P. Zhu, Z . Chen, J. Huang, Q. Sun, M. Feng, P. Chien, Q. Yang, and T. Cai, ”Preparation of phosphate glass composite and their prosthetic applications,” 1.Non-Cryst. Solids, 52, 503-510 (1982). E Pernot, P. Baldet, F. Bonnel, J. Zarzycki, and P. Rabischong, “Development of phosphate glass-ceramics for bone implants,” Ceram. Int., 9, 127-131 (1983). S. Kihara, A. Watanabe, and Y. Abe, “Calcium phosphate glass-ceramic crown prepared by lost-wax technique,” Am. Ceram. SOC. Comm., 67, C100-C101 (1984). W. Holand, W. Vogel, K. Naumann, and J. Gummel, ”Interface reactions between machinable bioactive glass-ceramics and bone,” J. Biomed. Mater. Res., 19, 303-312 (1985). T. Nakamura, T. Yamamuro, S. Higashi, T. Kakubo, and S. Itoo, ‘A new glass-ceramic for bone replacement: evaluation of its bonding to bone tissue,” J. Biomed. Mater. Res., 19, 685-698 (1985). H.V. Denissen, K. de Groot, P.C. Makkes, A. van den Hooff, and P. J. Klopper, ”Tissue response to dense apatite implants in rats,” J. Biomed. Mater. Res., 14, 713-721 (1980). I? Ducheyne and K. de Groot, “In vivo surface activity of a hydroxyapatite alveolar bone substitute,” 1. Biomed. Mater. Res., 15, 441-445 (1981). C. P. A.T. Klein, A. A. Driessen, K. de Groot, and A. van den Hooff, “Biodegradation behaviour of various calcium phosphate materials in bone tissue,“ J. Biomed. Mater. Res., 17, 769-784 (1983). C.P.A.T. Klein, A.A. Driessen, and K. de Groot, “Relationship between the degradation behaviour of calcium phosphate ceramics and their physical-chemical characteristics and ultrastructural geometry,” Biomaterials, 5, 157-160 (1984). A. Krajewski, A. Ravaglioli, L. Riva di Sanseverino, F. Marchetti, and G. Monticelli, ”The behaviour of apatite-based ceramics in relation to the critical 1150-1250°C temperature range,” Biomaterials, 5, 105-108 (1984). K. Kondo, M. Okuyama, H. Ogawa, Y. Shibata, and Y. Abe, ”Preparation of high-strength apatite ceramics,” Am. Ceram. SOC. Comm., 67, C222-C223 (1984). C.A. van Blitterswijk, J. J. Grote, W. Kuypers, C. J.G. Blok-van Hoek, and W. Th. Daems, “Bioreactions at the tissue/hydroxyapatite interface,” Biomaterials, 6, 243-251 (1985). P. Ducheyne, “Bioceramics: material characteristics versus in vivo behavior,” 1.Biomed. Mat. Res.: Appl. Biomat., 21 A2, 219-236 (1987). A. Krajewski, A. Ravaglioli, R. Mongiorgi, and A. Moroni, “Mineralization and calcium fixation within a porous apatitic ceramic material after implantation in the femur of rabbits,” 1. Biomed. Mater. Res., 22, 445-457 (1988). R. Z. LeGeros, J. R. Parsons, G. Daculsi, F. Driessens, D. Lee, S.T. Liu, S. Metsger, D. Peterson, and M. Walker, ”Significance of the porosity and physical chemistry of calcium phosphate ceramics,” Ann. N Y Acad. Sci., 523, 268-271 (1988).
CALCIUM PHOSPHATE MATERIALS CONTAINING ALUMINA 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.
35. 36. 37.
38. 39.
37
G. Daculsi, R. Z. LeGeros, and D. Mitre, ”Crystal dissolution of biological and ceramic apatites,” Calcif. Tiss. Int., 45, 95-103 (1989). B. A. Blencke, E. Pfeil, and H. H. Kas, “Implant aus glass-keramik in der knochenchirurgie (tier-experimentelle untersuchungen),” Langenbechs Arch. Chir. Suppl. Forum, 107-110 (1973). M. Jarcho, J. F. Kay, K. I. Gunmaer, R. H. Doremus, and H. P. Drobeck, ”Tissue, cellular and subcellular events at a bone-ceramic hydroxyapatite interface,” 1. Bioeng., 1, 79-92 (1977). M. Ogino, F. Ohuchi, and L. L. Hench, “Compositional dependence of the formation of calcium phosphate films on bioglass,” J. Biomed. Muter. Res., 14, 55-64 (1980). U. Gross and V. Strunz, ”The interface of various glasses and glass ceramics with a bony implantation bed,” J. Biomed. Muter. Res., 19, 251-271 (1985). B. A. Blencke, H. Bromer, and K. K. Deutscher, ”Compatibility and long-term stability of glass-ceramic implants,” J. Biomed. Muter. Res., 12, 307-316 (1978). T. Yuan, Z. Niu, M. Chou, H. Wang, D. Lin, X. Guo, Y. Jia, S. Chang, and J. Wang, ”Femoral components (cup) made of glass-ceramics,” J. Non-Cryst. Solids, 52, 487-496 (1982). G. Berger and V. Atzrodt, “In vitro characterization of bioactivity of glassy or glass-crystalline implant materials using Auger electron spectroscopy,” Phys. Stat. Sol., a(85), K9-Kl3 (1984). J.G. J. Peelen, B.V. Rejda, and K. de Groot, “Preparation and properties of sintered hydroxylapatite,” Ceram. Int., 4, 71-74 (1978). E. P. Egan, Jr. and Z.T. Wakefield, ”Thermodynamic properties of calcium metaphosphate, 10 to 1400”K,” J. Am. Chem. SOC., 78, 4245-4249 (1956). S. P. S. Porto and R. S. Krishnan, “Raman effect of corundum,”.J. Chem. Phys., 47, 1009-1012 (1967). D.C. O‘Shea, M. L. Bartlett, and R. A. Young, “Compositional analysis of apatites with laser-Raman spectroscopy: (OH, F, Cl)apatites,” Archs. Oral. Biol., 19, 995-1006 (1974). D. G. A. Nelson and 8.E. Williamson, ”Low-temperature laser Raman spectroscopy of synthetic carbonated apatites and dental enamel,” Aust. J. Chem., 35, 715-727 (1982). G. Daculsi, I. Orly, M. Gregoire, M. Heughebaert, D. J. Hartmann, and B. Kerebel, “Cell interactions with mixed calcium phosphate (TCP and HAP) and alumina solid phases: an ultrastructural study,” in Biological and Biomechanical Performance of Biomaterials, P. Christel, A. Meunier, and A. J.C. Lee (eds.), Elsevier Science Publishers B.V., Amsterdam, 1986, pp. 337-342. N.C. Blumenthal, A. S. Posner, V. Cosma, and U. Gross, “The effect of glass-ceramic bone implant materials on the in vitro formation of hydroxyapatite,” J. Biomed. Muter. Res., 22, 1033-1041 (1988). A. Bertoluzza, M. A. Battaglia, R. Simoni, and D. A. Long, “Vibrational spectroscopic studies of a new class of oligophosphate glasses of biological interest-Part 2,” J. Raman Spectr., 14, 178-183 (1983). A. Bertoluzza, C. Fagnano, A. Marinangeli, R. Simoni, A. Tinti, and M. A. Morelli, “Vibrational (Raman-laser and infrared) study on Ca(P03)2-A1203glasses,” in Biomaterials and Clinical Applications, A. Pizzoferrato, P.G. Marchetti, A. Ravaglioli, and A. J.C. Lee (eds.), Elsevier Science Publishers B.V., Amsterdam, 1987, pp. 531-516. H. Fukui, Y. Taki, and Y. Abe, “Implantation of new calcium phosphate glass-ceramics,” J. Dent. Res., 56, 1260 (1977). L. L. Hench and H. A. Paschall, ”Direct chemical bond of bioactive glass-ceramic materials to bone and muscle,” J. Biomed. Muter. Res. Symp., 4, 25-42 (1973).
BERTOLUZZA ET AL.
38 40. 41.
U. M. Gross and V. Strunz, “The anchoring of glass ceramics of different solubility in the femur of the rat,” ]. Biomed. Muter. Res., 14,607-618 (1980). P. Ducheyne, “Bioglass coatings and bioglass composites as implant materials,” 1. Biomed. Muter. Res., 19, 273-291 (1985).
Received July 5, 1989 Accepted April 24, 1990
Centro Studi Interfacoltli sulla Spettroscopia Raman, *Sezione di Chimica e Propedeutica Biochimica, Dipartimento di Biochimica, via Selmi 2, 40126 Bologna, #Dipartimento di Chimica “G. Ciamician: via Selmi 2, 40126 Bologna M. Morocutti, V. Ottani, and A. Ruggeri istituto di Anatomia Umana Normale, via Irnerio 48, 40226 Bologna, Italy Variable alumina quantities were added to two types of calcium phosphate materials -hydroxyapatite ceramics with Ca/P = 1.67 and calcium metaphosphate glass with Ca/P = 1-in order to increase their mechanical properties. Raman spectroscopy shows that alumina interacts with the phosphate group of these materials, while thermomechanical analysis shows that their elastic modulus has a value similar to that of bone. Histological sections demonstrate that the surface in close contact with ceramic materials
shows a good integration between bone and biomaterial. All ceramic specimens are penetrated by well-stained, presumably glycoprotein, matrix, that consistently forms a thin network in close contact with the implant. After 6 weeks bone growing on both ceramic and glass shows signs of maturation with a lamellar structure and an apparently normal mineralization. In the case of the glass, inside this newly formed bone, were often observed two layers, the internal one showing a well defined lamellar structure.
INTRODUCTION
Over the last several years efforts have been made to develop bone implant materials composed of calcium phosphate compounds because their chemical composition is closely allied to the mineral component of bone. Implant materials composed of calcium phosphate display excellent biocompatibility under many conditions and different forms of these implant materials have been studied in animals and The most widely clinically applied calcium phosphate ceramics are composed of either hydroxyapatite Ca,(PO,),(OH) (HA) or P-tricalciumphosphate Ca3(P04)2 (p-TCP) both with more or less successful results. Among materials containing calcium phosphate, but also with other components in certain quantity, are certain glasses and glass-ceramics.3,5’o ‘To whom correspondence should be addressed. Journal of Biomedical Materials Research, Vol. 25, 23-38 (1991) 0 1991 John Wiley & Sons, Inc. CCC 0021-9304/91/010023-16$04.00
24
BERTOLUZZA ET AL.
With regard to HA, depending on the function required by the implant site, there are several different forms of dense and porous hydroxyapatite available for bone implantation, each with characteristics that recently have been extensively st~died’~~~~’-” and which depend on the way the ceramic was prepared. A particular property of some calcium phosphate materials is their biodegradability. The factors which determine bioresorbability of ceramics are still controversial. However, it is generally thought that chemical composition, crystallographic structure, density, and porosity (in particular microporosity) can be identified as main parameter^.'",'^-'^,^^,^' Because of their apparent ability to bond directly to bone, these materials are called bioactive materials (i.e., hydroxyapatite, P-tricalciumphosphate, bioglasses, and glass-ceramics). Although not yet fully understood, the existence of this bond, and to some extent its appearance and composition, has been documented by numerous investigators using a variety of materials and animal models?,9.11,22-24 Despite their useful properties like biocompatibility and bone bonding, the main restriction of some calcium phosphate implant materials is their poor mechanical properties. Many investigations have been carried out in order to induce reinforcement of mechanical properties; some of these adding various oxides, like alumina, obtaining results not always e f f e ~ t i v e . ’ ~ ” * ~ ~ ~ ~ ~ The aim of the present study is to investigate the effect of the addition of alumina to hydroxyapatite and calcium metaphosphate glass and to determine bone tissue interaction when implanted in the tibiae of rats. MATERIALS A N D METHODS
Calcium hydroxyapatite (HA) powder with the composition CaS(PO&OH, supplied by Merck Co., was used for the preparation of ceramic materials. X-ray analysis on this product is described in the l i t e r a t ~ r e ’ ~ ”and ~ ~ shows ’~,~~ peaks of the hydroxylapatite structure. The powder was compressed uniaxially at 300 MN/m2 and sintered at 1200°C for 1 h. Then the ceramics were ground and newly sintered twice to improve the ductility of the material and also to obtain a greater homogeneity of the samples. Samples containing alumina (Erba product) were prepared in the same way. The porosity of the ceramics was calculated from the weight and volume of sintered samples. Total porosity ranged between 35% (pure HA) and 45% (ceramics containing alumina). Calcium metaphosphate glass, Ca(P03)2,was obtained by melting the crystalline metaphosphate P-Ca(P03)2at 1300°C for 1 h in a platinum crucible and then quickly cooling it. /3-Ca(P03)2was prepared by dehydration of monocalcium phosphate monohydrate (Fluka product), according to the 1iteratu1-e.~’ Glasses containing alumina were made by means of the same thermal process starting from monocalcium phosphate monohydrate (Ca(H2P04)2 . HzO) and alumina (A1203)so as to obtain the required ratio.
CALCIUM PHOSPHATE MATERIALS CONTAINING ALUMINA
25
The composition of the samples is listed in Table I. Ceramic (0.8 X 0.5 x 1.5 mm) and glass (dia. 0.5-0.8; 1 mm length) samples were implanted respectively into the right and the left tibia of the same Wistar rat (200 g average weight). Three animals for every two specimens were operated reaching a total of 15. Rats were anesthetized by intraperitoneal injections of sodium pentobarbital. The incision sites were shaved and cleaned with ethanol. Holes were drilled employing a diamond burr (dia. 0.8 mm) continuously cooled with physiological saline. After careful removal of bone debris, implants were inserted (press-fitted) and the wound closed by suture. Control holes were prepared the same way, but no implant was inserted. Animals were sacrificed 3, 4, and 6 weeks after operation. The segments of the tibiae were excised and immediately immersed in a 10% buffered formalin solution for histological processing. Because of the low solubility of the glasses into the decalcifying solution employed, it was necessary to remove them from the implant sites, and this was done after fixation to ensure a better sectioning. Hydroxyapatite implants were dissolved by demineralization into a solution of sodium citrate and formic acid, dehydrated in alcools and embedded in paraffin. Longitudinal and transversal histological sections were stained with hematoxylin-eosin or Azan Mallory stains. Raman spectra of glass and ceramics before and after implant into rat tibiae were recorded using a Jasco R 300 spectrometer and 600 mW of 488.0 nm radiation from a Spectra Physics argon ion laser. Dynamical load thermomechanical measurements (TMA) were carried out by means of a Mettler TA 3000 system equipped with a TMA 40 accessory at 37°C. Scanning electron microscopy (SEM) was carried out using a Phillips 501 electron microscope: one sample of every kind of material was mounted using colloidal silver and coated in argon atmosphere with a gold and palladium layer. TABLE I Ca/P and P/Al Ratio and Corresponding A1203Weight Percent of Vitreous and Ceramic Materials Studied Samples
Ca/P 1 1 1 1 1 1.67 1.67 1.67 1.67 1.67
u = vitreous, c = ceramic materials. "Sample 5 is not well vitrified.
PfA1
AlzOa weight% (w/w)
2.51 4.90 6.85 9.33 1.49 2.95 4.18 5.73
BERTOLUZZA ET AL.
26 RESULTS
Vibrational Raman spectroscopy Table I1 shows spectroscopical modifications of hydroxyapatite-alumina ceramics with various P/A1 ratio. As the alumina content in the ceramic increases, new bands at about 520,555,630, and 1090 cm-' appear in the spectra, and their intensity increases as alumina content does. These new bands are not present either in the alumina or hydroxyapatite Raman spectra. Table I11 shows spectroscopical modifications of calcium aluminophosphate glasses with various P/A1 ratio. As the alumina content increases, an increase in the frequency of the strong Raman band at about 690 cm-' in the calcium metaphosphate glass is observed. This band rises to 711 cm-' for the glass with 5/1 P/A1 ratio. On the contrary, there is a frequency decrease, less evident, of the strongest band at 1178 cm-' for calcium metaphosphate glass. This band is centered at about 1167 cm-l for the glass with the higher alumina content. A band practically absent in the spectrum of calcium metaphosphate glass appears in the spectra of calcium aluminophosphate glasses at about 1025 cm-' and augments in intensity increasing the alumina content of the glass. Thermomechanical analysis (TMA) Figure 1 shows TMA measurements performed on some glasses and ceramics and on rat tibiae (measured exactly in the same site in which the implant was performed). It is evident from the figure that glass samples have a TABLE I1 Raman Bands of Ceramic Materials HA 380 (39) 417 (100)
580 (10) 645 (30) 750 (36)
427 (14) 450 (sh) 578 (10) 590 (10) 604 (6) 947 (sh) 960 (100) 1025 (5) 1045 (8) 1075 (4) -
P/Al
=
2011
408 (sh) 430 (14) 450 (sh) 550 (2) 580 (10) 590 (10) 608 (7) 630 (sh) 947 (sh) 960 (100) 1025 (5) 1045 (8) 1075 (4) -
P/A1 = 10/1
410 (6) 430 (14) 450 (sh) 525 (sh) 552 (2) 580 (10) 592 (11) 610 (7) 630 (sh) -
947 (sh) 960 (100) 1025 (5) 1045 (7) 1075 (5) -
410 (6) 430 (14) 450 (sh) 525 (sh) 555 (2) 580 (10) 592 (11) 610 (7) 630 (4) 947 (sh) 960 (100) 1025 (sh) 1045 (7) 1075 (5) 1090 (sh)
sh = shoulder; numbers in parentheses are the intensities of the bands.
410 (11) 430 (14) 450 (sh) 520 (sh) 555 (2) 580 (10) 592 (11) 610 (8) 630 (4) 947 (sh) 960 (100) 1025 (sh) 1045 (7) 1075 (5) 1090 (sh)
CALCIUM PHOSPHATE MATERIALS CONTAINING ALUMINA
27
TABLE 111 Raman Bands of Glass Materials Al2O3
Ca(PO&
P/AI = 2011
P/AI = 1011
P/AI = 711
P/Al = 511
-
327 (12)
335 (16)
340 (19)
345 (22)
345 (25)
380 (39)
-
417 (100) 580 (10) -
395 (10)
-
-
-
-
525 (5)
-
600 (2) -
-
-
397 (sh)
395 (13)
-
525 (6)
525 (6)
-
-
605 (3)
610 (3)
-
612 (5)
612 (7) 711 (75)
-
-
698 (75)
705 (75)
795 (sh) 935 (1) 1023 (3) 1083 (6) 1178 (100) 1277 (15) 1342 (sh)
795 (sh) 935 (sh) 1025 (8) 1085 (sh) 1176 (100) 1266 (19) 1330 (sh)
797 (sh) 935 (sh) 1025 (14) 1100 (sh) 1171 (100) 1265 (20) 1330 (sh)
797 (sh) 935 (sh) 1025 (18) 1105 (sh) 1170 (100) 1265 (20) 1327 (sh)
-
-
-
-
525 (8)
-
698 (69)
-
-
397 (sh)
525 (7)
693 (51)
750 (36) -
395 (14)
-
645 (30)
-
-
-
6POz 6POP
-
-
V,POP 798 (sh) 935 ishi 1025 (28)
j va,pop
sh = shoulder; numbers in parentheses are the intensities of the bands.
I
I
1
0
0.1
0.2 PROeE
0
DISPLACEMENT
0.1
(*%)
Figure 1. Dynamic load thermomechanical analysis, corrected for the blank, of: (a) A calcium aluminophosphate glass with P/A1 = 10/1 (sample 3), A ceramic with P/Al = 10/1 (sample C), 0 rat tibiae bone; (b) A calcium aluminophosphate glasses with P/Al = 7/1 (sample 4), A P/A1 = 10/1 (sample 3) and calcium phosphate glass (sample 1).
BERTOLUZZA ET AL.
28
relatively higher compressive strength than ceramic samples. Moreover, TMA measurements performed on glass samples (1,3, and 4)and on ceramics (A, C, and E) reveal some differences in elastic behavior as a function of the composition (percent of alumina) of either glass or ceramic materials. In particular, the addition of alumina increases the compressive strength. The Young modulus calculated from these measurements is of the order of 1.5.10' Pa for ceramics, 2.5.10* Pa for glasses and 6.107 Pa for rat bone.
Scanning electron microscopy Before implant the microstructure of HA blocks consists of regular and rounded particles connected by grain boundaries resulting from the sintering procedure. The microporosity measured by weight and volume of sintered samples ranges from 35% for pure HA to 45% for samples containing the greater amount of alumina (Fig. 2). Glass blocks show an almost fully smooth surface and an amorphous microstructure. Incidental fractures show a sharp-edged shape typical of glass.
Implantation
Ceramic Histological sections do not show any inflammatory reactions from the first time period on three weeks studied. On decalcified sections HA speci-
Figure 2. SEM showing the HA-A1203material (sample C) porosity. Original magnification X4750.
CALCIUM PHOSPHATE MATERIALS CONTAINING ALUMINA
29
mens show a "vacuolized" appearance presumably caused by the material grains dissolved with decalcifying solution. Specimens with alumina addition have a more microgranular pattern and a higher density with respect to pure HA (Fig. 3). Four weeks after implant both HA and HA with A1,03 specimens are still completely surrounded by primary bone laid also down on medullary surface. The narrow layer of primary bone with woven fibers still shows a clear separation from the preexisting lamellar bone (Fig. 4). Some concentric lamellae are also visible. The surface in close contact with the implant shows a good integration between bone and biomaterial. All the specimens are penetrated by well-stained, presumably glycoprotein, matrix that all along forms a thin network in close contact with the implant. At 6 weeks we can still find primary bone but its appearance is becoming more lamellar and more integrated with preexisting bone mostly in subendosteal and periosteal areas. Specimens with higher content of A1203 show an increased thickness of the stained matrix at the surface of the implant. In some cases inside larger
Figure 3. Longitudinal histological section of HA specimen after 6 weeks showing a vacuolized aspect. Implant is completely surrounded by primary bone. Azan-Mallory staining. I = implant. Original magnification ~ 3 0 .
30
BERTOLUZZA ET AL.
Figure 4. Longitudinal histological section of HA-A1203 material (sample E) showing a microgranular pattern. The specimen is surrounded by new bone also on medullary surface. 4 weeks after implant. AzanMallory staining. I = implant. Original magnification x8.
lacunae at the implant surface some thin connective branches are observable (Fig. 5).
Glass No significant differences were seen in comparison with the interactions of different compositional glasses with the surrounding tissue. Four weeks after the implant the specimens were already surrounded by newly formed bone apparently without any connective tissue at their interface. Only in a few samples a thin layer of stainable material, occasionally containing cell debris, was observable. It is also apparent that tissue adjacent to where the implant was removed was predominantly osteoid and was also surrounding the medullary canal. The boundary between old and newly formed bone was always clearly visible because of differences in stainability and lamellar orientation (Fig. 6). After 6 weeks bone growing on both glass with and without A1203addition showed signs of maturation which was represented by a prevalent substitution of primary bone with lamellar structure. Inside this newly formed bone it was frequently possible to notice the presence of two layers. On the one hand, bone directly bordering the external implant surface had more matured attaining a concentrical lamellar pattern. On the other hand the outer
CALCIUM PHOSPHATE MATERIALS CONTAINING ALUMINA
31
Figure 5. Longitudinal histological section of HA-A1203material (sample D) 6 weeks after operation showing connective branches at the implant surface. Azan-Mallory staining. I = implant. Original magnification ~ 1 8 5 .
layer of newly formed bone, not yet completely lamellar, still showed a boundary with the old bone, but a better integration was visible (Fig. 7). In control animals in almost all of the cases newly formed bone attains an aspect very similar to that of a normal secondary bone within 6 weeks.
DISCUSSION
TMA measurements show that the elastic modulus of our glass and ceramic materials is quite similar to that of rat tibiae. In both materials the increase of the Young modulus for addition of alumina suggests the possibility of changing the elastic characteristics of the material as a function of the requirements of different implant sites.
BERTOLUZZA ET AL.
32
Figure 6. Longitudinal histological section of glass (sample 2) 4 weeks after operation showing the clear separation between old and newly formed bone. (-) Azan-Mallory staining. Original magnification X50.
Ceramic
Concerning hydroxyapatite-alumina ceramics, Table I1 evidences that the band at 410 cm-' for ceramics with higher alumina content can derive from the more intense band at 417 cm-' in the alumina spectrum.31In the same way the new band at 630 cm-' can be related to the band at 645 cm-' in the alumina Raman spectrum. The new bands at 520 and 555 cm-' can be attributed to modifications in the structure of phosphate groups of HA induced by the interaction with A1203,as already reported by other authors for v4 stretching modes of phosphate groups in substituted apatite^.^^,^^ Moreover, the two bands attributed to phosphate groups at 578 and 590 cm-I only partially change their frequency owing to this interaction. A new weak band at about 1090 cm-', not present in the spectrum of alumina, is evident in ceramics containing more alumina. This can be related to the 1075cm-' band of hydroxyapatite. These spectral modifications show that the sintering process of ceramics has led to the synthesis of materials where the Raman bands of AlzO, and of the phosphate group are lightly modified. These changes can be related to a chemical interaction that occurred during the sintering process. Density measurements show that HA with alumina has a greater microporosity than normal HA. Moreover, SEM micrographs of the surface of HA with alumina samples show an apparent increase of microporosity, with a greater number of small pores.
CALCIUM PHOSPHATE MATERIALS CONTAINING ALUMINA
33
Figure 7. Transversal histological section of glass (sample 4) showing the presence of two layers: the one closer to the implant has a concentrical ) staining. Original magnification X 52. lamellar pattern. (4Azan-Mallory
The presence of A13+ in the sintered material leads to a delay in the grain growthz5and to a decrease in the sizes of the pores. From decalcified histological sections, the material of samples B, C, D, and E appears more microgranular, probably owing to alumina residual, after the material has been pervaded and stabilized by the organic component of the matrix. All the HA-implanted materials show a good tissue integration as demonstrated by the absence of any inflammatory or toxic reactions of the tissue in the vicinity of implants. Histological sections in contact with the implant surface typically show bone growing in direct contact with the implant material without any intervening fibrous tissue capsule. HA materials completely surrounded on all sides allow the formation of a good attachment between implant and bone. During histological processing HA is firmly adherent to the bone so that an eventual breakage preferably takes place between new and old bone rather than between new bone and implant. Raman spectra of ceramic materials after implant show only some stretching bands which can be attributed to organic matter, probably caused by organic phase of bone. The mineral phase of bone, nevertheless, cannot be observed because its bands are coincident with those of HA. The osteointegration and the progressive maturation of the bone surrounding the implant material is probably better for samples containing alu-
34
BERTOLUZZA ET AL.
mina than for pure HA. Samples used in our study are well pervaded by a substance which may be glycoprotein containing stainable material which on the surface becomes fairly thick and is almost organized into some superficial lacunae on small connective branches. This behavior contrasts with the hypothesis that addition of this oxide to Ca/P biomaterials normally has a negative effect on their bio~ompatibility?~,~~ The discrepancies in the results obtained are probably related to a difference in the percent of alumina employed or otherwise give rise to some doubts about the inhibitory effects of alumina.35 Glass
As shown in Table 111, addition of basic oxides to metaphosphate glasses produces a demolition of the polymeric structure and the formation of oligomers characterized by an intensity increase of the stretching vibration of PO:- terminal groups at about 1020-1040 cm-' and a decrease of 1170- and 690-cm-' bands of the polymeric structure.36 The spectroscopical trend for calcium metaphosphate glasses containing A1203is similar to that of oligophosphates, but the new band attributable to PO:- groups terminal of the chain is much less intense and localized at a lower wavenumber (1025 cm-I). On the contrary, the lowering of the 690 cm-' band, relative to POP groups, is of the same order as that observed for calcium oligophosphate glasses. This behavior can be interpreted considering that the presence in the chains of alumina breaks the vibrational coupling of POP groups, but the lower presence of PO?- groups suggests a partial covalent interaction A1-0 with respect to the prevailingly ionic interaction of calcium.37 Histological observations on rat tibiae where vitreous materials have been implanted show, in any case, a good response of the host tissue, and indicate an excellent tissue integration. calcium With regard to this item, as already reported by other metaphosphate gives good results, and also in agreement with Zhu et a1.6our glasses containing alumina are well tolerated by the tissue. Seemingly contradictory results were obtained by Daculsi and his when they showed that alumina added to Ca/P materials reduces their biocompatibility. Indeed their experimental conditions differ from ours, as their material was implanted in connective tissue rather than in bone. Moreover their ultrastructural observations are restricted to cell reactivity while our data include the environment of the host tissue in its globality. Our data also show the absence of signs of any inflammatory reaction, the absence of a fibrous capsule and, indeed, a generalized formation of new bone which extends on all sides of the implant, including the medullary one. Already 4 weeks after implant, at the interface with the glass, a bone lamellar structure is observable and this demonstrates an advanced stage of bone remodelling. One may suggest that the glassy material, possibly because of its elastic modulus (which is similar to that of bone), is effective in the sec-
CALCIUM PHOSPHATE MATERIALS CONTAINING ALUMINA
35
ondary bone formation. This behavior is also observable in samples containing 10% in weight of alumina. An addition of this kind of oxide would have an inhibitory effect on mineralization. In fact, these oxides are used to stabilize and delay the release of ions from vitreous material^?^*^^^^^,^^,^^,^^ The fact that alumina in our samples appears to have no influence seems to derive from the weak effect of this oxide or from the characteristic structure of the glass itself. Moreover, the absence of any visible acute inflammatory reaction should be evidence in favor of a surface controlled reaction of the vitreous material. The poor solubility of a vitreous material does not seem to be a sure parameter for the absence of bone since at the surface of the scarcely soluble material it is possible that after a long time an adsorption of proteins capable of promoting a deposition of a inorganic Ca/P layer could take place. In fact, the Raman spectrum of one of our glasses after implant revealed the presence of a weak band centered at about 960 cmP, which can be related to an apatitic phase on the surface of the glass. For other samples this band was not evident, probably because of the difficulties in obtaining the surface Raman spectra of small transparent materials. Maybe, the Ca/P rich layer is not by itself a sufficient reason for bone bonding under functionally loaded conditions. Possibly implant surface porosity or microstructure is more important than the effect of ions leached from the implant s ~ r f a c e . ~ ~ , ~ ’ Limited to 6 weeks, our samples do not show a great adhesion to the surrounding tissue and this behavior confirms the hypothesis that the presence of alumina in glass is detrimental to bone b ~ n d i n g . ~ ~However, , ~ ~ , ~ * , consider~’ ing the limited time of our experiment we cannot exclude that bone bonding takes place after a longer time, i.e., 12 weeks, as estimated elsewhere for nearly insoluble glass with addiction of a l ~ m i n a . ~ ~ , ~ ~ In conclusion, it has been observed that mechanical and elastic properties of calcium phosphate materials (i.e., ceramics and glasses) increase as the alumina content of the material increases. Regarding the materials with a hydroxyapatite structure there is a sintering process of alumina-phosphate evidenced by vibrational spectroscopy, and by Scanning Electron Microscopy. On the contrary, in calcium metaphosphate, alumina induces a breakage of the polyphosphate and formation of oligomers with shorter chains, where PO;- terminal groups are bonded with a partial covalent character to aluminium. As to the biocompatibility of calcium phosphate materials containing alumina, it has been observed that small quantities of alumina are well tolerated. Both implant materials are surrounded by newly formed bone that become progressively mature showing a good osteointegration in the case of ceramic material, and a formation of two bone layers in that of the glass. References 1. M. Jarcho, ”Calcium phosphate ceramics as hard tissue prosthetics,” Clin. Orthop. Rel. Res., 157, 259-278 (1981). 2. T. Han, F. A. Carranza, and E. B. Kenney, ”Calcium phosphate ceramics in dentistry: a review of the literature,” J. West. SOC.Periodont. Periodont. Ah., 32, 88-108 (1984).
BERTOLUZZA ET AL.
36 3. 4.
5. 6. 7.
8.
9.
10. 11. 12. 13. 14.
15.
16.
17. 18. 19.
20.
F. Pernot, R. Rogier, J. Zarzycki, F. Bonnel, and P. Baldet, “Les biomateriaux vitreux et vitroceramiques en chirurgie orthopedique et OSseuse,” Bull. SOC. Ckim. Fr., 4, 519-522 (1985). K. de Groot, ”Degradable ceramics,” in Biocompatibility of Clinical Implant Materials, Vol. I, D. F. Williams, (ed.), CRC Press, Boca Raton, 1981, pp. 199-222. L. L. Hench and A. E. Clark, ’Adhesion to bone,” in Biocompatibility of Orthopaedic Implants, Vol. 11, D. F. Williams (ed.), CRC Press, Boca Raton, 1982, pp. 129-170. P. Zhu, Z . Chen, J. Huang, Q. Sun, M. Feng, P. Chien, Q. Yang, and T. Cai, ”Preparation of phosphate glass composite and their prosthetic applications,” 1.Non-Cryst. Solids, 52, 503-510 (1982). E Pernot, P. Baldet, F. Bonnel, J. Zarzycki, and P. Rabischong, “Development of phosphate glass-ceramics for bone implants,” Ceram. Int., 9, 127-131 (1983). S. Kihara, A. Watanabe, and Y. Abe, “Calcium phosphate glass-ceramic crown prepared by lost-wax technique,” Am. Ceram. SOC. Comm., 67, C100-C101 (1984). W. Holand, W. Vogel, K. Naumann, and J. Gummel, ”Interface reactions between machinable bioactive glass-ceramics and bone,” J. Biomed. Mater. Res., 19, 303-312 (1985). T. Nakamura, T. Yamamuro, S. Higashi, T. Kakubo, and S. Itoo, ‘A new glass-ceramic for bone replacement: evaluation of its bonding to bone tissue,” J. Biomed. Mater. Res., 19, 685-698 (1985). H.V. Denissen, K. de Groot, P.C. Makkes, A. van den Hooff, and P. J. Klopper, ”Tissue response to dense apatite implants in rats,” J. Biomed. Mater. Res., 14, 713-721 (1980). I? Ducheyne and K. de Groot, “In vivo surface activity of a hydroxyapatite alveolar bone substitute,” 1. Biomed. Mater. Res., 15, 441-445 (1981). C. P. A.T. Klein, A. A. Driessen, K. de Groot, and A. van den Hooff, “Biodegradation behaviour of various calcium phosphate materials in bone tissue,“ J. Biomed. Mater. Res., 17, 769-784 (1983). C.P.A.T. Klein, A.A. Driessen, and K. de Groot, “Relationship between the degradation behaviour of calcium phosphate ceramics and their physical-chemical characteristics and ultrastructural geometry,” Biomaterials, 5, 157-160 (1984). A. Krajewski, A. Ravaglioli, L. Riva di Sanseverino, F. Marchetti, and G. Monticelli, ”The behaviour of apatite-based ceramics in relation to the critical 1150-1250°C temperature range,” Biomaterials, 5, 105-108 (1984). K. Kondo, M. Okuyama, H. Ogawa, Y. Shibata, and Y. Abe, ”Preparation of high-strength apatite ceramics,” Am. Ceram. SOC. Comm., 67, C222-C223 (1984). C.A. van Blitterswijk, J. J. Grote, W. Kuypers, C. J.G. Blok-van Hoek, and W. Th. Daems, “Bioreactions at the tissue/hydroxyapatite interface,” Biomaterials, 6, 243-251 (1985). P. Ducheyne, “Bioceramics: material characteristics versus in vivo behavior,” 1.Biomed. Mat. Res.: Appl. Biomat., 21 A2, 219-236 (1987). A. Krajewski, A. Ravaglioli, R. Mongiorgi, and A. Moroni, “Mineralization and calcium fixation within a porous apatitic ceramic material after implantation in the femur of rabbits,” 1. Biomed. Mater. Res., 22, 445-457 (1988). R. Z. LeGeros, J. R. Parsons, G. Daculsi, F. Driessens, D. Lee, S.T. Liu, S. Metsger, D. Peterson, and M. Walker, ”Significance of the porosity and physical chemistry of calcium phosphate ceramics,” Ann. N Y Acad. Sci., 523, 268-271 (1988).
CALCIUM PHOSPHATE MATERIALS CONTAINING ALUMINA 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.
35. 36. 37.
38. 39.
37
G. Daculsi, R. Z. LeGeros, and D. Mitre, ”Crystal dissolution of biological and ceramic apatites,” Calcif. Tiss. Int., 45, 95-103 (1989). B. A. Blencke, E. Pfeil, and H. H. Kas, “Implant aus glass-keramik in der knochenchirurgie (tier-experimentelle untersuchungen),” Langenbechs Arch. Chir. Suppl. Forum, 107-110 (1973). M. Jarcho, J. F. Kay, K. I. Gunmaer, R. H. Doremus, and H. P. Drobeck, ”Tissue, cellular and subcellular events at a bone-ceramic hydroxyapatite interface,” 1. Bioeng., 1, 79-92 (1977). M. Ogino, F. Ohuchi, and L. L. Hench, “Compositional dependence of the formation of calcium phosphate films on bioglass,” J. Biomed. Muter. Res., 14, 55-64 (1980). U. Gross and V. Strunz, ”The interface of various glasses and glass ceramics with a bony implantation bed,” J. Biomed. Muter. Res., 19, 251-271 (1985). B. A. Blencke, H. Bromer, and K. K. Deutscher, ”Compatibility and long-term stability of glass-ceramic implants,” J. Biomed. Muter. Res., 12, 307-316 (1978). T. Yuan, Z. Niu, M. Chou, H. Wang, D. Lin, X. Guo, Y. Jia, S. Chang, and J. Wang, ”Femoral components (cup) made of glass-ceramics,” J. Non-Cryst. Solids, 52, 487-496 (1982). G. Berger and V. Atzrodt, “In vitro characterization of bioactivity of glassy or glass-crystalline implant materials using Auger electron spectroscopy,” Phys. Stat. Sol., a(85), K9-Kl3 (1984). J.G. J. Peelen, B.V. Rejda, and K. de Groot, “Preparation and properties of sintered hydroxylapatite,” Ceram. Int., 4, 71-74 (1978). E. P. Egan, Jr. and Z.T. Wakefield, ”Thermodynamic properties of calcium metaphosphate, 10 to 1400”K,” J. Am. Chem. SOC., 78, 4245-4249 (1956). S. P. S. Porto and R. S. Krishnan, “Raman effect of corundum,”.J. Chem. Phys., 47, 1009-1012 (1967). D.C. O‘Shea, M. L. Bartlett, and R. A. Young, “Compositional analysis of apatites with laser-Raman spectroscopy: (OH, F, Cl)apatites,” Archs. Oral. Biol., 19, 995-1006 (1974). D. G. A. Nelson and 8.E. Williamson, ”Low-temperature laser Raman spectroscopy of synthetic carbonated apatites and dental enamel,” Aust. J. Chem., 35, 715-727 (1982). G. Daculsi, I. Orly, M. Gregoire, M. Heughebaert, D. J. Hartmann, and B. Kerebel, “Cell interactions with mixed calcium phosphate (TCP and HAP) and alumina solid phases: an ultrastructural study,” in Biological and Biomechanical Performance of Biomaterials, P. Christel, A. Meunier, and A. J.C. Lee (eds.), Elsevier Science Publishers B.V., Amsterdam, 1986, pp. 337-342. N.C. Blumenthal, A. S. Posner, V. Cosma, and U. Gross, “The effect of glass-ceramic bone implant materials on the in vitro formation of hydroxyapatite,” J. Biomed. Muter. Res., 22, 1033-1041 (1988). A. Bertoluzza, M. A. Battaglia, R. Simoni, and D. A. Long, “Vibrational spectroscopic studies of a new class of oligophosphate glasses of biological interest-Part 2,” J. Raman Spectr., 14, 178-183 (1983). A. Bertoluzza, C. Fagnano, A. Marinangeli, R. Simoni, A. Tinti, and M. A. Morelli, “Vibrational (Raman-laser and infrared) study on Ca(P03)2-A1203glasses,” in Biomaterials and Clinical Applications, A. Pizzoferrato, P.G. Marchetti, A. Ravaglioli, and A. J.C. Lee (eds.), Elsevier Science Publishers B.V., Amsterdam, 1987, pp. 531-516. H. Fukui, Y. Taki, and Y. Abe, “Implantation of new calcium phosphate glass-ceramics,” J. Dent. Res., 56, 1260 (1977). L. L. Hench and H. A. Paschall, ”Direct chemical bond of bioactive glass-ceramic materials to bone and muscle,” J. Biomed. Muter. Res. Symp., 4, 25-42 (1973).
BERTOLUZZA ET AL.
38 40. 41.
U. M. Gross and V. Strunz, “The anchoring of glass ceramics of different solubility in the femur of the rat,” ]. Biomed. Muter. Res., 14,607-618 (1980). P. Ducheyne, “Bioglass coatings and bioglass composites as implant materials,” 1. Biomed. Muter. Res., 19, 273-291 (1985).
Received July 5, 1989 Accepted April 24, 1990
