J O U R N A L O F M AT E R I A L S S C I E N C E : M AT E R I A L S I N M E D I C I N E 1 3 ( 2 0 0 2 ) 1 1 9 9 ± 1 2 0 6
`
`A comparative study on the in vivo behavior of
`hydroxyapatite and silicon substituted
`hydroxyapatite granules
`
`N. PATEL*, S. M. BEST, W. BONFIELD
`Department of Materials Science and Metallurgy, University of Cambridge,
`New Museums Site, Pembroke Street, Cambridge, CB2 3QZ, UK
`E-mail: np239@cam.ac.uk
`
`I. R. GIBSON, K. A. HING
`IRC Biomedical Materials, Queen Mary University of London, Mile End Road, London,
`E1 4NS, UK
`
`E. DAMIEN, P. A. REVELL
`IRC Biomedical Materials, Royal Free and University College Medical School,
`Royal Free Campus, Rowland Hill Street, London, NW3 2QG, UK
`
`Phase pure hydroxyapatite (HA) and a 0.8 wt % silicon substituted hydroxyapatite (SiHA)
`were prepared by aqueous precipitation methods. Both HA and SiHA were processed into
`granules 0.5±1.0 mm in diameter and sintered at 1200 C for 2 h. The sintered granules
`underwent full structural characterization, prior to implantation into the femoral condyle of
`New Zealand White rabbits for a period of 23 days. The results show that both the HA and
`SiHA granules were well accepted by the host tissue, with no presence of any in¯ammatory
`cells. New bone formation was observed directly on the surfaces and in the spaces between
`both HA and SiHA granular implants. The quantitative histomorphometry results indicate
`that the percentage of bone ingrowth for SiHA (37.5% + 5.9) was signi®cantly greater than
`that for phase pure HA (22.0% + 6.5), in addition the percentage of bone/implant coverage
`was signi®cantly greater for SiHA (59.8% + 7.3) compared to HA (47.1% + 3.6). These
`®ndings indicate that the early in vivo bioactivity of hydroxyapatite was signi®cantly
`improved with the incorporation of silicate ions into the HA structure, making SiHA an
`attractive alternative to conventional HA materials for use as bone substitute ceramics.
`# 2002 Kluwer Academic Publishers
`
`1. Introduction
`Hydroxyapatite [Ca10(PO4)6(OH)2, HA] is well estab-
`lished as a synthetic material for bone replacement due to
`its chemical similarities to the inorganic component of
`bone and tooth. However, the mineral phase of bone is
`not hydroxyapatite, but can be described as a multi-
`substituted calcium-phosphate apatite. The type and
`amount of ionic substitution in the apatite phase varies
`from the wt % level (e.g. 2±8 wt % CO3) to the ppm-ppb
`level (e.g. Mg2‡ or Sr2‡ ) [1, 2]. The role of many of
`these ionic species in hard tissues is not fully understood
`owing to the dif®culties encountered in monitoring and
`quantifying the amounts of ionic content in bone mineral,
`which vary in composition consequent upon dietary
`alterations and physiological and pathological causes [3].
`However, it is accepted that these different ions play a
`major role in the biochemistry of bone.
`in skeletal
`Silicon is believed to be essential
`development,
`the ®rst
`indications of a physiological
`
`role for silicon were determined by Carlisle [4, 5] who
`reported that silicon was involved in the early stage of
`bone calci®cation. Similarly studies by Schwarz and
`Milne [6] have shown that silicon de®ciency in rats
`resulted in skull deformations, with the cranial bones
`appearing more ¯atter than normal. These early ®ndings
`suggest a relationship between silicon intake and bone
`mineralization.
`More recent evidence of the role of silicon in bone
`metabolism can be found in the use of silicon rich
`materials (bioactive glasses and glass ceramics) that
`contain high levels of SiO2 (30±50 wt %) [7, 8]. It has
`been proposed that the bioactivity of these SiO2-rich
`glass formulations is related to the role of SiO2 or silicon
`on their surface reactions, and therefore in their in vivo
`and in vitro bioactivity [9]. The initial cellular and ionic
`processes which occur at the surface of bioactive glasses
`and glass ceramics allow the subsequent crystallization
`of apatite crystals, cell adhesion and collagen formation.
`
`* Author to whom all correspondence should be addressed.
`
`0957±4530 # 2002 Kluwer Academic Publishers
`
`(cid:20)(cid:3)(cid:82)(cid:73)(cid:3)(cid:27)
`
`MILLENIUM EXHIBIT 2022
`Baxter Healthcare Corp. et al. v. Millenium Biologix, LLC
` IPR2013-00582, -00590
`
`1199
`
`

`

`These reactions in glasses are very rapid and have been
`reported to occur within minutes of implantation [10],
`and it is these rapid responses which are thought to make
`these materials very bioactive.
`Since SiO2 or silicon has been shown to enhance the
`bioactivity of glass-based materials, silicon may also
`enhance the bioactivity of HA with the incorporation of
`silicon into the HA structure. HA has the advantage over
`bioactive glasses and glass ceramics in that
`it
`is
`chemically similar to the inorganic component of bone.
`Therefore a potential method for
`improving the
`bioactivity of hydroxyapatite is the incorporation of
`silicon into the hydroxyapatite lattice. Many attempts
`have been made to prepare silicon substituted hydro-
`xyapatites by a variety of synthesis methods, however
`these have resulted in either silicon containing apatites
`with undesirable secondary phases [11, 12] or
`the
`additional substitution (besides silicon) of ionic groups
`into the HA lattice [13]. These secondary phases and
`additional substitutions make it dif®cult to determine the
`true role of silicon on the in vitro and in vivo bioactivity
`of silicon substituted hydroxyapatite.
`A recent study by Gibson et al. [14] described the
`preparation of a silicon substituted hydroxyapatite by an
`aqueous precipitation method. The resulting silicon
`substituted hydroxyapatite
`contained approximately
`0.4 wt % silicon, with no indication of any secondary
`phases such as calcium oxide or tricalcium phosphate. In
`addition, the authors reported that the in vitro bioactivity
`was substantially enhanced with the incorporation of
`silicon into the HA lattice [15]. However, in vitro tests
`provide limited information, as they do not simulate the
`true physiological in vivo behavior of the material. To
`date there have been no reports which compare the in
`vivo bioactivity of SiHA with HA. The aims of this
`present study were to synthesize and prepare granules of
`phase pure HA and SiHA, and to assess the effect of
`silicon substitution on the in vivo bioactivity of HA.
`
`2. Materials and methods
`2.1. Sample preparation
`Stoichiometric hydroxyapatite was prepared by an
`aqueous precipitation reaction between calcium hydro-
`xide (Ca(OH)2) and orthophosphoric (H3PO4) acid
`solution based on the methods described elsewhere
`[16±18]. The precipitation reaction was performed at
`room temperature and the pH was maintained at
`approximately 10.5 with the addition of ammonia
`solution. After complete mixing of the reactants, the
`suspension was aged at room temperature overnight.
`A 0.8 wt % silicon substituted HA was synthesized in a
`similar manner
`to HA according to the methods
`described by Gibson et al. [14]. The quantities of
`reactants were calculated on the basis that the number of
`moles of H3PO4 in phase pure HA would be equivalent to
`the number of moles of (H3PO4 ‡ silicon acetate) in the
`
`silicon substituted hydroxyaptite, with the number of
`moles of Ca(OH)2 kept constant at 0.5. Both sets of ®lter
`cakes were processed into green granules (0.5±1 mm) by
`partial grinding and mechanical sieving. The granules
`were subsequently sintered to 1200 C for 2 h in air
`1200
`
`(Carbolite RHF 1600); the heating rate was 2.5 C/min
`and the cooling rate to room temperature was 10 C/min.
`
`2.2. Chemical analysis
`The phase composition of both HA and SiHA granules
`were analyzed by X-ray diffraction (XRD) using a
`Siemens D5000 diffractometer. Data were collected over
`the 2y range 25±40 C with a step size of 0.02 and a
`count time of 2.5 s. Identi®cation of the phases was
`achieved by comparing the diffraction patterns of HA
`and SiHA with ICDD (JCPDS) standards [19]. The
`calcium/phosphorus (Ca/P) molar ratios of HA and SiHA
`were determined by a direct measurement from the X-ray
`¯uorescence (XRF) spectroscopy results. The carbonate
`contents of the sintered granules were measured by C±H±
`N analysis (Medac Ltd. UK). Fourier transform infrared
`(FTIR) spectra were obtained using a Nicolet 800
`spectrometer, at a resolution of 4 cm 1, averaging
`100 scans.
`
`2.3. Physical characterization
`The bulk density for both sets of sintered granules was
`determined using a Micromeritics Accupyc 1330
`pycnometer. In addition the packing density of the
`granules was determined using a Micromeritics Geopyc
`1360 instrument, by measuring the mass of granules
`required to ®ll a container of speci®ed volume. The
`speci®c surface area of the granules was determined by
`the Brunauer±Emmett±Teller (BET) method using a
`Micromeritics Gemini II 2370 surface area analyzer.
`Scanning electron microscopy (SEM) of the sintered
`granules was performed using a ®eld emission JEOL
`F6300 SEM, which provided information about
`the
`morphology and size of the granules.
`
`2.4. Implantation procedure
`The guidelines for the care and use of laboratory animals
`(Animals
`(Scienti®c Procedures) Act 1986) were
`observed throughout the implantation procedures. Six
`month old female New Zealand White rabbits (average
`weight & 3 kg) were used for implantation. Prior to
`surgery granules of sintered HA and SiHA were sterilized
`at 150 C for 4 h. The implants were placed bilaterally
`into the femoral condyle (Fig. 1). Defects approximately
`4 mm in diameter and 7 mm in depth were drilled into the
`patellar groove of the femoral condyle using a saline
`cooled diamond tip trephine. The sterile granules were
`press-®tted into the defect prior to suturing. The animals
`were injected subcutaneously with ¯uorochrome labels
`according to the protocol described in Table I, and were
`sacri®ced 23 days after surgery.
`
`2.5. Histological evaluation
`After sacri®ce the implant/bone were ®xed in formalde-
`hyde solution and dehydrated using various grades of
`alcohol, prior to embedding in PMMA resin. Sectioning
`of the specimen blocks was performed parallel to the
`longitudinal axis of the tibia using a diamond band saw
`and an Exakt system for grinding and polishing thin
`
`2 of 8
`
`

`

`Figure 1 Position of hydroxyapatite (HA) or silicon substituted hydroxyapatite (SiHA) granular implants, (a) schematic diagram identifying the key
`points, and (b) photograph of the implant site.
`
`T A B L E I Fluorochrome labeling protocol
`
`Fluorochrome
`label
`
`Dosage
`(mg kg 1)
`
`Time of label
`administration
`(days after surgery)
`
`Appearance under
`UV light with N.2.1.
`Leica ®lter cube
`
`Appearance under
`UV light with RGB
`Leica ®lter cube
`
`Tetracycline
`Calcein green
`Alizarin red
`Tetracycline
`
`0.5
`0.5
`0.5
`0.5
`
`0
`7
`14
`21
`
`Animals were sacri®ced 23 days after surgery.
`
`sections (Exakt Corp., Hamburg, Germany). Sections
`& 25 mm thick were stained with toluidine blue for
`histological examination and histomorphometry using a
`Leitz
`bright®eld
`transmitted
`optical microscope.
`Unstained sections also & 25 mm thick were examined
`using a Leica DMRXB UV light optical microscope to
`detect ¯uorescent labels.
`
`2.6. Histomorphometry
`Histomorphometry was performed on micrographs of
`sections of HA and SiHA taken at magni®cations
`of 6 25. A series of images were obtained for each
`section. These images were stitched together to represent
`a collage of the whole section. Histomorphometry was
`performed on each section using linear intercept and
`point counting methods in order to build up a map of
`bone ingrowth and bone/implant coverage within each
`section. The percentage of bone ingrowth was measured
`for each section using a 42-point Weibel grid [20], a hit
`
`over an area of bone, as shown in Fig. 2, the percentage
`
`for bone ingrowth …HBI† was scored when a point fell
`of bone ingrowth per Weibel grid …AGBI† was calculated
`ingrowth for the whole section …ASBI† was determined as
`a mean of the number …NW† of Weibel grids used per
`
`according to Equation 1. The absolute percentage of bone
`
`section according to Equation 2.
`Piˆ 42
`iˆ 1 HBIi
`42
`Piˆ NW
`iˆ 1 AGBIi
`NW
`
`AGBI ˆ
`
`AGBI ˆ
`
`6100
`
`…1†
`
`…2†
`
`Similarly the coverage of bone ingrowth on the implant
`surfaces (bone coverage) was measured by placing Merz
`grids [21] over the section, as shown in Fig. 3. A hit for
`
`bone coverage …HBC† was scored when a line intersected
`space …HIS† was
`
`a bone/implant interface, similarly a hit for implant/
`scored where a line
`trabecular
`intersected an implant/trabecular space interface,
`the
`
`Orange
`Green
`Red
`Orange
`
`Yellow
`Green
`Red
`Yellow
`
`calculated according to Equation 3. The absolute
`
`percentage of bone coverage per Merz grid …GBC† was
`percentage of bone coverage for the section …SBC† was
`calculated as a mean of the number …NM† of Merz grids
`
`used per section, according to Equation 4.
`Piˆ x
`iˆ 1 HBCi
`Pi ˆ y
`jˆ 1
`
`6100
`
`HISi
`
`GBC ˆ
`
`SBC ˆ
`
`Piˆ NW
`iˆ 1 GBCi
`NM
`
`…3†
`
`…4†
`
`The ¯uorochrome labels were used to assess the bone
`mineral apposition rates (MAR) of HA and SiHA
`sections. Fluorochrome labels are bound to sites of
`active bone deposition, shortly after administration. This
`enables identi®cation of bone deposited at different time
`points. The mineral apposition rates were calculated by
`
`Figure 2 Measurement of bone ingrowth with a Weibel grid. A hit for
`bone ingrowth …HBI† was scored when a point fell over an area of bone.
`1201
`
`3 of 8
`
`

`

`Figure 3 Measurement of bone coverage with a Merz grid. A hit for
`bone coverage …HBC† was scored when a line intercepted a bone/implant
`interface. Similarly a hit for implant/trabecular space …HIS† was scored
`
`when a line intercepted an implant/trabecular space interface.
`
`measuring the mean distance between the three different
`¯uorescent labels, and dividing these distances by the
`difference in the time-points at which the labels were
`administrated.
`
`3. Results
`3.1. Chemical analysis
`The Ca/P ratio and the principal elements detected by
`XRF for HA and SiHA are shown in Table II. The Ca/P
`ratio of HA was found to be equivalent to the theoretical
`value of phase pure HA (1.67). The Ca/P molar ratio of
`SiHA (1.74) was much higher
`than the value for
`stoichiometric HA. This increase in Ca/P ratio for
`SiHA is due to the substitution of silicate ions for the
`phosphate site in HA, in order to accommodate silicate
`ions, the phosphorus content was decreased. However, if
`the molar ratio for SiHA is measured as Ca/(P‡ Si) ratio,
`then the value of 1.67 is the same as the molar ratio of
`HA. The amounts of silicon (wt %) as measured by XRF
`are also shown in Table II. The amount of silicon in the
`SiHA sample (0.78 wt %) was very close to the expected
`value (0.8 wt %), the HA granules contained undetect-
`able amounts of silicon as expected. Furthermore, the
`carbonate contents measured by C±H±N analysis
`indicate that for both HA and SiHA sintered granules,
`carbonate content was undetected. The X-ray diffraction
`patterns of sintered HA and SiHA granules are shown in
`Fig. 4. The XRD patterns for the two materials appeared
`to be similar, the incorporation of silicon into the HA
`lattice did not have a direct affect on the phase
`composition, with no indication of secondary phases
`such as calcium oxide (CaO) or tricalcium phosphate
`(TCP). The diffraction peaks matched the ICDD
`(JCPDS) standards for HA. Fig. 5 show the FTIR spectra
`
`sintered (1200 C; 2 h)
`Figure 4 X-ray diffraction patterns of
`hydroxyapatite (HA) and silicon substituted hydroxyapatite (SiHA)
`granules. No secondary phases such as Ca3(PO4)2 (TCP) or CaO were
`detected.
`
`Figure 5 Fourier
`transform infrared (FTIR) spectra of sintered
`(1200 C; 2 h) hydroxyapatite (HA) and silicon substituted hydroxy-
`apatite (SiHA) granules.
`
`of sintered HA and SiHA and are similar to those
`observed by Gibson et al. [14] with the most notable
`effect of silicon substitution on the FTIR spectra of
`hydroxyapatite being the changes in the PO4 bands
`between 800 and 1100 cm 1 and 500±700 cm 1.
`
`3.2. Physical characterization
`The bulk density, packing density and speci®c surface
`areas of both sets of granules are listed in Table III. The
`bulk density for HA and SiHA was found to be 97±99%
`of the theoretical value (3:156 g cm 3). The packing
`density was similar for both HA and SiHA granules
`indicating that the same mass of granules for both cases
`would be required to ®ll defects of known volume. The
`speci®c surface areas of the sintered granules were also
`similar. Fig. 6 shows SEM images of sintered HA and
`SiHA granules, both sets of granules were similar in size
`and morphology.
`
`3.3. Histological evaluation
`Extensive osseointegration was observed in both HA and
`SiHA implants with seams of osteoblasts depositing bone
`
`T A B L E I I Chemical analysis of sintered HA and SiHA granules using X-ray ¯uorescence (XRF) spectroscopy, showing the measured silicon
`content compared to the calculated value, and the Ca/(P‡ Si) molar ratio. Carbonate content was measured from the C±H±N elemental analysis
`Ca/(P‡ Si) molar ratio
`
`wt % C (measured)
`
`Sample
`
`wt % Si (measured)
`
`Ca/P molar ratio
`
`HA
`SiHA
`
`1202
`
`5 0.01
`0.78
`
`1.67
`1.74
`
`1.67
`1.67
`
`5 0.01
`5 0.01
`
`4 of 8
`
`

`

`T A B L E I I I Bulk density, packing density and speci®c surface area measurements of HA and SiHA granules sintered at 1200 C for 2 h. (All
`densities were quoted as a percentage of the theoretical density for HA; 3:156 g cm 3:)
`
`Sample
`
`HA
`SiHA
`
`Bulk density (%)
`
`Packing density (%)
`
`Speci®c surface area (sq mg 1)
`
`97.2
`99.7
`
`43.8
`42.6
`
`2.36
`2.63
`
`directly on the surface of the granular implants, as shown
`in Fig. 7. High magni®cation micrographs, Fig. 7(b) and
`(d), indicate that new bone had formed directly on the
`surface and within the spaces between the granules of
`HA and SiHA implants. There was no evidence of any
`in¯ammatory cells on the surface or between the granular
`implants and ®brous encapsulation was not observed.
`Cells such as osteoblasts and osteocytes were observed in
`close proximity to the implant surfaces. Active areas of
`bone deposition and resorption was observed occurring
`within all
`implants. Fluorescent microscopy, Fig. 8,
`indicates the dominant presence of layers of both alizarin
`red and calcein green bone marking labels in the bone
`tissue around the HA and SiHA granules. Both calcein
`green and alizarin red labels were observed continuously,
`in some cases in direct contact to the surface of the HA or
`SiHA granules and in some cases several microns away
`from the granule surface. In some regions (within both
`HA and SiHA implants), the presence of diffuse calcein
`green and alizarin red labels were observed, indicating
`woven bone, which had been rapidly formed. The
`intensity of
`the tetracycline label was very weak,
`possibly due to bleaching under UV light during
`¯uorescent microscopy. The primary direction of bone
`ingrowth seemed to occur from the deep end toward the
`super®cial end of the defect, with additional bone
`ingrowth from the walls of the defect.
`
`3.4. Histomorphometry
`A quantitative assessment of the area of bone ingrowth,
`Fig. 9, indicates that the absolute percentage of bone
`(37.52% + 5.87) was
`ingrowth for SiHA implants
`signi®cantly greater than that for HA (21.99% + 6.85)
`granular implants. In addition the area of bone coverage
`or bone on-growth (a measurement of the amount of
`contact between the implant and bone), as shown in Fig.
`10 was signi®cantly greater for SiHA (59.77% + 7.34)
`implants compared to HA (47.08% + 3.60%) implants.
`The bone mineral apposition rates for all implants were
`
`implantation.
`calculated during weeks 1±2 after
`Determination of the apposition rates between weeks 2
`and 3 was not possible, due to the weak intensity of the
`tetracycline label which was administrated 3 weeks after
`surgery. The bone mineral apposition rates, Fig. 11, for
`SiHA (3.91 + 0.27 mm/day) implants were again sig-
`ni®cantly greater
`than that
`for phase pure HA
`(2.53 + 0.29 mm/day).
`
`4. Discussion
`Recent studies have highlighted how the in vitro
`bioactivity of hydroxyapatite can be enhanced with the
`incorporation of small levels of physiologically relevant
`ions such as silicate [15], carbonate [22] or magnesium
`[23] ions into the hydroxyapatite lattice. In vivo studies
`on these synthetic substituted hydroxyapatites are,
`however limited in number and quality, one of the
`reasons is that the physical properties of the substituted
`hydroxyapatite implants used, e.g. shape, size, and
`density may be very different from the control, normally
`HA implants.
`In addition,
`the synthesis methods
`employed to prepare substituted hydroxyapatites, often
`result in the presence of secondary phases or the added
`substitution of unwanted secondary ions, which can have
`a detrimental affect on the in vivo properties of these
`substituted apatites.
`In this study granules (0.5±1 mm diameter) of phase
`pure hydroxyapatite and 0.8 wt % silicon substituted
`hydroxyapatite were prepared with similar physical
`properties. The synthesis route used to prepare a
`0.8 wt % silicon substituted HA was successful and the
`desired amount of silicon in the sintered granules was
`achieved, as demonstrated by the XRF results. The XRF
`results also reveal that the number of phosphate groups in
`the SiHA granules was reduced, as indicated by the
`higher Ca/P molar ratio compared to HA granules.
`the Ca/(P‡ Si) molar ratio for SiHA was
`However,
`calculated as 1.67, which is equivalent to that for phase
`pure HA. The XRF results suggest that the silicon was
`
`Figure 6 Scanning electron micrographs showing that the sintered: (a) hydroxyapatite (HA) and (b) silicon substituted hydroxyapatite (SiHA) were
`similar in size and morphology.
`
`1203
`
`5 of 8
`
`

`

`Figure 7 Histological appearance of (a) bone ingrowth within hydroxyapatite (HA) granular implants, (b) direct bone apposition onto the surface of
`HA implants, (c) bone ingrowth within silicon substituted hydroxyapatite (SiHA) granular implants and (d) direct bone apposition on the surface of a
`SiHA granule.
`
`incorporated into the HA lattice where it occupied the
```phosphorus site'', according to the reaction mechanism
`described elsewhere [14]. Further evidence of
`the
`incorporation of
`silicon into the HA lattice was
`demonstrated by the XRD data shown in Fig. 4. No
`
`secondary phases were observed for the SiHA granules,
`and the diffraction pattern for SiHA was identical to that
`of phase pure HA granules. If the silicate ions had not
`substituted for the phosphate site in HA, the resulting
`material
`that would have been produced would be
`
`Figure 8 Fluorochrome labeled bone within (a) hydroxyapatite (HA) implants (image acquired using a Leica N.2.1. ®lter cube), (b) silicon
`substituted hydroxyapatite (SiHA) implants (image acquired using a Leica RGB ®lter cube) and (c) HA implants (RGB ¯uorescent image
`superimposed onto a bright-®eld transmitted light image) demonstrating lamellar bone deposited at 1±2 (calcein green; CG) and 2±3 (alizarin red;
`AR) weeks after implantation.
`
`1204
`
`6 of 8
`
`

`

`Figure 9 Percentage of bone ingrowth within hydroxyapatite (HA) and
`silicon substituted hydroxyapatite (SiHA) implants …p 5 0.05†.
`
`calcium rich resulting in a secondary calcium oxide
`phase. In this study, the XRD data revealed no secondary
`phases for both sintered HA and SiHA granules. FTIR
`spectroscopy, Fig. 5, of HA and SiHA revealed
`differences in the phosphate n3 and n1 region (1100±
`900 cm 1), which was due to the substitution of silicon
`(or silicate) into some of the phosphorous (or phosphate)
`sites in HA. Furthermore no structural carbonate groups
`were observed by FTIR or C±H±N elemental analysis for
`the sintered HA and SiHA granules, demonstrating that
`the substitution was purely silicon (or silicate) for
`phosphorus (or phosphate).
`Granules of HA and SiHA prepared in this study where
`similar in size and morphology, as indicated by the SEM
`images shown in Fig. 6. In addition the speci®c surface
`area, bulk density ( & 97±98% of the theoretical) and the
`packing density ( & 43%) of both HA and SiHA granules
`are similar (Table III) indicating that the same mass of
`granules were required to ®ll a defect of known volume.
`The importance of granule size, shape, bulk density and
`packing density of bioceramics when implanted in an
`osseous environment has been reported by Oonishi et al.
`[24] where they reported signi®cant differences between
`osseointegration with granule size and density. Similarly
`Ikeda et al.
`[25]
`showed that
`the rate of bone
`regeneration for porous HA and apatite-wollastonite
`glass ceramics was dependent on granule size and
`porosity. In this study we prepared granules of HA and
`SiHA with similar size/shape, speci®c surface area, bulk
`density and packing density, prior to the application of an
`in vivo study.
`The in vivo results indicate bone ingrowth into the
`spaces between both HA and SiHA granules, with direct
`bone contact between the granules and the surrounding
`bone. Both HA and SiHA implants were well accepted in
`the host tissue, with no evidence of in¯ammatory cells.
`No mobility between the granular implants and the bone
`was observed during the sectioning procedure, indicating
`good ®xation of the implants. The absolute percentage of
`bone ingrowth, coverage and bone mineral apposition
`rate for SiHA implants was signi®cantly greater than that
`for HA at 23 days in vivo. These ®ndings support other in
`vitro and in vivo studies which have established that
`silicon plays an active role in bone formation and
`In vitro studies on 0.8 and
`calci®cation [4, 5, 15].
`1.6 wt % silicon substituted hydroxyaptite by Gibson et
`al. [15] have suggested that the acceleration of bone
`
`Figure 10 Percentage of bone coverage on the surface of hydro-
`xyapatite (HA) and silicon substituted hydroxyapatite (SiHA) implants
`…p 5 0.05†.
`
`apposition for silicon substituted hydroxyapatite may
`partly result from an up-regulation in osteoblast cell
`metabolism, however the mechanism by which this
`occurs has to still
`to be resolved. Alternatively the
`physiological degradation properties of SiHA may differ
`from HA. De Bruijn et al. [26] reported that the in vivo
`solution-mediated degradation of a calcium phosphate
`was an important factor for bone/implant integration at
`early time points after implantation, whereby bone tissue
`formed more rapidly on the surface of calcium
`phosphates which resorbed.
`Further studies are required to determine the long-term
`response of a range of silicon substituted hydroxyapatite
`implants
`compared to phase pure hydroxyapatite
`implants. This study shows clearly, however, that the
`biological activity/response of hydroxyapatite ceramics
`was signi®cantly enhanced by the substitution of low
`levels of silicate ions into the HA lattice.
`
`5. Conclusions
`Granules of phase pure hydroxyapatite and 0.8 wt %
`silicon substituted hydroxyapatite with similar size,
`morphology, bulk density, packing density and speci®c
`surface area were prepared. The in vivo results from this
`study indicate that the bioactivity of hydroxyapatite was
`signi®cantly enhanced with the addition of silicate ions
`into the hydroxyapatite structure. These ®ndings suggest
`that silicon substituted hydroxyapatite is an improved
`alternative to phase pure hydroxyapatite for biomedical
`applications.
`
`Figure 11 Apposition rates of bone ingrowth for hydroxyapatite (HA)
`and silicon substituted hydroxyapatite (SiHA) implants …p 5 0.05†.
`
`1205
`
`7 of 8
`
`

`

`14.
`
`15.
`
`20.
`
`I . R . G I B S O N , S . M . B E S T and W. B O N F I E L D , J. Biomed.
`Mater. Res. 44 (1999) 422.
`I . R . G I B S O N , J . H U A N G , S . M . B E S T and W. B O N F I E L D , in
```Proceedings of the 12th International Symposium on Ceramics
`in Medicine, Nara, Japan, October 1999'', edited by H. Oghushi,
`G. W. Hastings and T. Yoshikawa (World Scienti®c Publishing
`Co., Singapure, 1999) p. 191.
`16. M . A K AO , H . AO K I and K . K AT O , J. Mater. Sci. 28 (1981)
`809.
`17. M . J A R C H O , C . H . B O L E N , M . B . T H O M A S , J . B O B I C K , J . F.
`K AY and R . H . D O R E M U S , ibid. 11 (1976) 2027.
`I . R . G I B S O N , S . K E , S . M . B E S T and W.
`18. N . P AT E L ,
`B O N F I E L D , J. Mater. Sci. Mater. in Med. 12 (2001) 181.
`19. PDF card no. 09-0432, ICDD, Newton Square, Pennsylvania,
`USA.
`E . R . W E I B E L and H . E . E L I A S , in ``Quantitative methods in
`morphology'' (Springer-Verlag, Berlin, 1967).
`21. W. A . M E R Z and R . K . S C H E N K , Acta. Anat. 76 (1970) 1.
`22. K . A . H I N G , J . C . M E R R Y, I . R . G I B S O N , L . D I S I LV I O , S . M .
`B E S T and W. B O N F I E L D ,
`in ``Proceedings of
`the 12th
`International Symposium on Ceramics in Medicine, Nara, Japan,
`October 1999'', edited by H. Oghushi, G. W. Hastings and
`T. Yoshikawa (World Scienti®c Publishing Co. Singapure)
`p. 195.
`T. J . W E B S T E R , C . E R G U N , R . H . D O R E M U S and R . B I Z I O S ,
`J. Biomed. Mater. Res. 59 (2002) 312.
`24. H . O O N I S H I , L . L . H E N C H , J . W I L S O N , F. S U G I H A R A , E .
`T S U J I , S . K U S H I TA N I and H . I WA K I , ibid. 44 (1999) 31.
`25. N . I K E D A , K . K AWA N A B E and T. N A K A M U R A , Biomaterials
`20 (1999) 1087.
`J . D . D E B R U I J N , Y. P. B OV E L L , J . E . D AV I E S and C . A . VA N
`B L I T T E R S W I J K , J. Biomed. Mater. Res. 28 (1994) 105.
`
`23.
`
`26.
`
`Received 24 May
`and accepted 29 May 2002
`
`Acknowledgments
`The support of the Engineering and Physical Sciences
`Research Council for their funding of a studentship to N.
`Patel. The authors would also like to thank Mr Tom
`MacInnes for his assistance in processing sections for
`histology. Dr Roger Brooks and Dr Susan Clarke are
`acknowledged for
`their assistance in analyzing the
`histomorphometry data.
`
`References
`1. N . C . B L U M E N T H A L , F. B E T T S and A . S . P O S N E R , Calcif.
`Tissue Res. 18 (1975) 81.
`2. W. D . A R M S T R O N G and L . S I N G E R , Clinical Orthopaedics 38
`(1965) 175.
`3. A . E . S O B E L , M . R O C K E N M AC H E R and B . K R A M E R , J. Biol.
`Chem. 235 (1960) 2502.
`E . M . CA R L I S L E , Science 167 (1970) 179.
`4.
`E . M . CA R L I S L E , Calc. Tissue Int. 33 (1981) 27.
`5.
`6. K . S C H WA R Z and D . B . M I L N E , Nature 239 (1972) 333.
`L . L . H E N C H , J. Phys. 43 (1982) 625.
`7.
`T. KO K U B O , S . I T O , Z . T. H U A N G , T. H AY A S H I , S . S A K K A ,
`8.
`T. K I T S U G I and T. Y A M A M U R O , J. Biomed. Mater. Res. 24
`(1990) 331.
`L . L . H E N C H and A . E . C L A R K E , in ``Biocompatibility of
`Orthopaedic Implants'', edited by D. F. Williams (CRC Press,
`Inc., 1982) Chapter 6.
`L . L . H E N C H , J. Am. Ceram. Soc. 74 (1991) 1487.
`10.
`11. A . J . R U Y S , J. Aust. Ceram. Soc. 29 (1993) 71.
`12. K . S U G I Y A M A , T. S U Z U K I and T. S AT O H , J. Antibact.
`Antifung. Agents 23 (1995) 67.
`13. Y. TA N I Z AWA and T. S U Z U K I , Phosphorus Res. Bull. 4 (1994)
`83.
`
`9.
`
`1206
`
`8 of 8
`
`