Graft Copolymers of Methyl Methacrylate and Poly([R]-3-hydroxybutyrate) Macromonomers as Candidates for Inclusion inAcrylic Bone Cement Formulations: Compression Testing

Sophie Nguyen, Robert H. Marchessault

Department of Chemistry, McGill University, Montreal, Quebec, Canada

Received 10 November 2004; revised 26 May 2005; accepted 1 June 2005Published online 3 October 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jbm.b.30430

Abstract: Graft copolymers of methyl methacrylate and biodegradable, biocompatible bac-terial poly([R]-3-hydroxybutyrate) (PHB) blocks were synthesized and evaluated as possibleconstituents in acrylic bone cements for use in orthopaedic applications. The copolymers wereproduced by conventional free radical copolymerization and incorporated in one commer-cially available acrylic bone cement brand, Antibiotic Simplex® (AKZ). Cements with for-mulations containing 6.7 and 13.5 wt % of PMMA-graft-PHB were prepared. The morphologyof the graft copolymer particles was suggested to influence the ability of the modified cementto be processed. Formulations containing more than about 20 wt % of the graft copolymerresulted in cement doughs that, both after first preparation and several hours later, wereeither sandy or soft spongy in texture and, thus, would be unacceptable for use in orthopaedicapplications. The morphologies of the powders and the volumetric porosity (p) and ultimatecompressive strength (UCS) of the cured cements were determined. Micro computed tomog-raphy showed that the cements presented average porosities of 13.5–16.9%. It was found that,while the powder particle shape and size for the experimental cements were markedlydifferent from those of AKZ, there was no significant difference in either p or UCS for thesecements. The latter was determined to be about 85 MPa for the modified cements and 84 MPafor Antibiotic Simplex. Furthermore, the UCS of all the cements exceeded the minimum levelfor acrylic bone cements, as stipulated by ASTM F-451. © 2005 Wiley Periodicals, Inc. J BiomedMater Res Part B: Appl Biomater 77B: 5–12, 2006

Keywords: graft copolymers; poly([R]-3-hydroxybutyrate); poly(methyl methacrylate);bone cements; compression

INTRODUCTION

Acrylic bone cements are materials based on poly(methylmethacrylate) (PMMA) and have been widely used in ortho-pedics for total joint (e.g. hip, shoulder, or knee) replacementsurgery1,2 since the pioneering work of Charnley and Smithin the early sixties.1–3 The cement is prepared in the surgeryroom just before use, by mixing a liquid monomer, mainlymethyl methacrylate (MMA), with a powder, mainly aPMMA-based polymer. MMA polymerizes around the swol-len polymer, forming a dough. In the interior of the bone,where a cavity has been created by the removal of bonemarrow and trabeculae, the cement dough is injected to fixthe prosthesis. The result is a nonadhesive mechanical inter-locking to the bone as the cement penetrates the intertrabe-cular spaces and conforms to the shape of the inner cortical

bone, as well as to the surface irregularities present in mostprostheses.4 In the 1980s, acrylic bone cements were alsointroduced to treat vertebral compression fractures caused byosteoporosis, skeletal metastases, and angiomas.5–8 One pro-cedure is vertebroplasty, which consists of a percutaneousinjection of a bone cement dough into a fractured vertebralbody to relieve pain and to prevent further compression andreduction in height.5–8

In total joint replacement, the cured acrylic cement allowsefficient primary fixation, supports body weight, and facili-tates transfer of service loads to the bone.9,10 However, onecommon complication is aseptic loosening of the implant,which occurs in about 10% of the patients undergoing jointreplacement. The consequence is a reduction of the jointreplacement lifetime to less than 10 years,9 compared with 15years in 90% of recently cemented arthroplasty operations.2

To counteract premature loosening, several strategies havebeen reported, one of them being the development of partiallydegradable cement.11–14 Classical PMMA-based bone ce-ments are inert, that is do not resorb or allow bone ingrowth.7

With partially degradable cements, it is hypothesized that

Correspondence to: R. H. Marchessault (e-mail: robert.marchessault@mcgill.ca)Contract grant sponsor: The Natural Sciences and Engineering Research CouncilContract grant sponsor: Labopharm Inc. provided financial support.

© 2005 Wiley Periodicals, Inc.

5

bone would grow around, as well as into, the cement in thespace left by the degraded material.15 This would result instronger fixation of the prosthesis in the bone cavity.11–15

Examples in the literature of bone cement formulationscontaining degradable polymeric constituents are relativelyscarce. They include biodegradable blends of corn starch andcellulose acetate, in combination with biostable PMMA andhydrophilic poly(acrylic acid) or poly(2-hydroxymethylmethacrylate).14–16 Such formulations, with hydroxyapatiteas a biocompatible and osteoconductive ceramic filler, werestudied by Espigares et al.,14 while Boesel and Reis used thecommercial bioactive glass (Bioglass® 45S5) as filler.16

Bone cement formulations using biodegradable crosslinkablepoly(propylene fumarate) (PPF) oligomers as the polymercomponent with calcium carbonate and tricalcium phosphateas fillers have been reported.17 Wise and coworkers haveproposed bioerodible semi-interpenetrating network alloysfor surgical plates and bone cements in the surgical repair oforthopaedic and maxillofacial fractures.18 These alloys werecomposed of polymers such as copolymers poly(lactic acid-co-glycolic acid), polydioxanone, or poly(�-caprolactone),which are biodegradable through hydrolysis, and PPFcrosslinked with vinyl pyrrolidone or MMA, to provide in-ternal reinforcement.18

Polyhydroxyalkanoates (PHAs) are bacterial aliphaticpolyesters that are biodegradable19–22 and biocompat-ible.20–22 Zinn et al. reported potential applications of PHAsin orthopaedics, such as scaffolds for cartilage engineering,bone graft substitutes, meniscus regeneration, and spinalcages22 and other internal fixation devices.22,23 Poly([R]-3-hydroxybutyrate) (PHB), the most studied member of thisfamily of microbial materials, naturally occurs in humanblood in its monomeric form, (R)-3-hydroxybutyric acid.20–22

It is also found as short chains (degree of polymerization lessthan 150) in prokaryotic and eukaryotic cells, in competentEscherichia coli, mammalian mitochondria, and in the humanblood stream, but in low concentrations.24,25 PHB slowlydegrades in vivo, through hydrolysis.20–22 However, unlikeother degradable polymers that degrade faster, such as poly(L-lactic acid), PHB features piezoelectric properties, whichcould promote bone growth.23,26 For example, compositescontaining 70 vol % of poly(3-hydroxybutyrate-co-7% hy-droxyvalerate) (PHB/V7 mol%) and 30 vol % of hydroxyapa-tite were studied as implants for fracture treatment in rabbitfemurs.23 The interface with bone showed marked bone in-growth in the absence of any adverse tissue reaction.23 Inter-face mechanical shear strength increased up to the first 8weeks as hard tissue bonded to the implant surface, anddecreased afterwards with PHB/V7 mol% degradation and soft-ening due to water uptake.23

Biocompatible and partially biodegradable bone cementformulations containing (a) PHB or PHV; (b) a hydrosolublepolymer (e.g. poly(acrylic acid), or poly(methacrylic acid));and (c) a nonhydrosoluble, nonhydrolysable one (e.g.PMMA) were proposed by Bunel and coworkers.11,12 In onecase, PHA was contained in the powder component, whereasthe other polymers were included either in the powder or the

liquid monomer.11 In another case, a chemically modifiedPHA was included in the liquid component as a crosslinkingmonomer, with a methacrylic or acrylic function at each endgroup.12 Jiang and Roby also presented partially as well ascompletely bioabsorbable acrylic bone cement formulations,in which the powder component contained PHB or PHV.13

Thus, copolymers of MMA and low molecular weightbacterial PHB, PMMA-graft-PHB, would be interesting can-didates as new polymeric materials for bone cement formu-lations featuring partial degradability and compatibility withthe other methacrylic components. The synthesis of suchcopolymers, involving the free radical copolymerization ofMMA with methacrylic macromonomers of PHB, PHB*, (DPca. 20) was reported earlier,27 and their formula is shown inFigure 1. The objective of the present study was two-fold.First, to determine the compositions of acrylic bone cementthat include this copolymer in its powder, and which maythen be processed to produce a dough of acceptable texture.Second, to establish whether the experimental formulationsso obtained are viable alternatives to a current commerciallyavailable acrylic bone cement brand (Antibiotic SurgicalSimplex, AKZ).

More specifically, in this study two different parts of thepowder component in the commercial kit were replaced byPMMA-graft-PHB, while the liquid component was directlyused from the kit. Since benzoyl peroxide (BPO), the initiatorof MMA polymerization, represented 1.6 wt % of the poly-mer content, it was added to the new powder component sothat it amounted to 1.6 wt % of the graft copolymers, and toa ratio of 0.028–0.029 g of total BPO (the one added and theone already contained in the commercial powder) per mL ofliquid monomer. The in vivo performance of the cementedarthroplasty is affected by the physical and mechanical prop-erties of the acrylic bone cement (among other factors). In thepresent work, the volumetric porosity and quasi-static com-pressive properties (ultimate strength, ultimate strain, andmodulus) of the PMMA-graft-PHB substituted AKZ cementwere determined as a first assessment of the cement’s poten-tial for use in orthopaedic applications. Porosity was mea-sured using microtomography, and discussed in terms ofmicroporosity, when the pore diameter is between 0.1 and 1mm, and macroporosity, when the pore diameter is over 1

Figure 1. Chemical formula of graft copolymers of MMA and lowmolecular weight bacterial PHB, PMMA-graft-PHB.

6 NGUYEN AND MARCHESSAULT

mm.2,4 Compressive properties were chosen as the first me-chanical properties to determine, since the cement transfersbody weight and service loads between the bone and theartificial joint.2

MATERIALS AND METHODS

Materials

Bone cement kits, Antibiotic Simplex®, or AKZ (StrykerCanada LP, Howmedica Osteonics, Guelph, ON) were storedat 3°C and used as received. BPO (Anachemia Chemicals)was purified by dissolution in chloroform (CHCl3; Fisher,HPLC grade), filtration, and precipitation by the addition ofHPLC grade methanol (CH3OH) with a volume ratio CHCl3:CH3OH of 1:2. The suspension was then filtered, and thesolid product was dried under vacuum and stored at 3°C. Forthe copolymerization, MMA (Aldrich, 99%) was stirred overcalcium hydride, distilled under reduced pressure, and storedat �20°C. 2,2�-Azobisisobutyronitrile (AIBN; BDH Chemi-cals Ltd) was recrystallized from methanol, and stored at 3°C.Anisole (Aldrich, anhydrous, 97%) was washed with a 2Msodium hydroxide aqueous solution, then with distilled water,and dried over calcium chloride. The solvent was then dis-tilled under reduced pressure from barium oxide, and storedat �20°C. For the recovery and purification of products,methanol and methylene chloride (both from Fisher, HPLCgrade) were not further purified. MMA (Aldrich, 99%, stabi-lized with 10–100 ppm of monomethyl ether hydroquinone)was used for the solubility tests of PHB*.

Preparation and Characterization of PMMA-graft-PHB

PMMA-graft-PHB was obtained from the free radical copo-lymerization of PHB*,27 (number-average molecularweight � 2010, polydispersity ratio � 1.2, and functional-ity � 100%) with MMA, in anisole at 70°C, using AIBN aspolymerization initiator. PHB* (11.3 g, 5.61 mmol) andMMA (11.6 g, 116 mmol) (4.6 mol % of PHB* in thecomonomer feed ratio) were placed in a round bottom flask,and the suspension was purged with argon (Ar). Anisole (179mL) was transferred under Ar to the reaction flask, which waspurged back with Ar. AIBN (0.0967 g, 0.589 mmol) wascharged in a flask with a cap equipped with a septum andpurged with argon for at least 30 min. Anisole (30 mL) wasthen transferred under argon to the vial containing the AIBN,which was stirred until complete dissolution, and purgedagain with Ar. This solution was transferred under Ar to thereaction flask, which was then placed in a 70°C heatedsilicone oil bath and kept there for 45 h under magneticstirring. The reaction flask was then removed from the hotbath, cooled down with cold tap water, and the product wasprecipitated out with methanol. The copolymer was purifiedby dissolution in methylene chloride and precipitation by theaddition of methanol.

Proton Nuclear Magnetic Resonance

1H NMR was used for the determination of the structure ofthe polymers, PMMA-graft-PHB and polymer in AKZ pow-der, as well as their comonomer molar composition. 1H NMRexperiments were performed on a Varian Unity 500 MHz atroom temperature, with the samples dissolved in deuteratedchloroform (CDCl3) and tetramethylsilane as internal stan-dard.

Gel Permeation Chromatography

Analyses were performed at room temperature with chloro-form as eluent, at a flow rate of 1.0 mL/min, using twoWaters Styragel columns HR3 and HR4 connected in series.PMMA standards were used for calibration, with a HewlettPackard refractive index HP 1047 RI detector.

Bone Cement Formulations, Specimen Preparation, andCharacterization of Cements

Bone Cement Formulations. Four different formula-tions, T100, T50, T10, and T25, were tested for their abilityto be processed. Three different formulations were finallychosen for porosity and compression testing and designatedF0, F10, and F20. The powder compositions for all formula-tions are shown in Table I. In all formulations, the liquidmonomer was the one from the AKZ kit. The proportions ofthe powder and liquid components were as follows: 4.0 g ofthe powder was mixed with 2.0 mL of the liquid monomer,the manufacturer’s recommended ratio being 2.18 g/mL.

In formulation F0, the powder was 100% from the sameAKZ kit, so the obtained samples were considered as refer-ences. Formulations F10 and F20 corresponded to 6.6–6.7 wt% and 13.5 wt % of graft copolymer in the resulting cement,respectively.

Bone Cement Mixing and Transfer. The powder compo-nent and the utensils used were kept at room temperature

TABLE I. Powder Composition for the Different FormulationsUsed, the Liquid Component Being the One from an AntibioticSimplex® Kit. In All Formulations, 4.0 g of Powder Was MixedWith 2.0 mL of Liquid.

CementBatch No.

Powder Composition (wt %)

AntibioticSimplex® PMMA-graft-PHB BPOa

T100 0 98.4 1.6T50 49.2 49.9 0.94T10 89.8 10.0 0.15T25 75.0 24.6 0.39F0 100 0 0F10 90.1 9.8 0.16F20 79.9 19.8 0.31

BPO, benzoyl peroxide; the values shown are the amount added to the commercialpowder from AKZ kit, along with the graft copolymers. The commercial powderalready contains 1.4% wt % of the total powder, corresponding to 1.6 wt % of thepolymer, and 0.029 g of BPO per mL of liquid monomer.

7COMPRESSION TESTING OF PHB ENRICHED ACRYLIC BONE CEMENTS

while the liquid monomer was chilled at �20°C, to improvethe workability of the cement and reduce its porosity.2 Themixing and the transfer of the bone cement were performed atroom temperature in a 30-mL regular taper tip poly(pro-pylene) syringe (McMaster-Carr). The syringe tip was cutback 1 cm, to obtain an orifice diameter of 2 mm, for a betterflow of the produced paste. Typically, the powder (4.0 g) wastransferred to the syringe barrel with the tip blocked by asmall pipe cleaner brush. In the case of F10 and F20, the threecomponents of the powder were added separately at roomtemperature. The liquid monomer (2.0 mL) was removedfrom the freezer, immediately added to the powder, and thereaction medium was mechanically stirred for 1 min at about115 rpm. The stirring device was a stainless steel twistedpaddle. The pipe brush was then removed, and the syringepiston was used to transfer the obtained paste to the mold.

Molding and Trimming. Solid cylindrical specimens forthe uniaxial compression testing were prepared in accordancewith ASTM F451–86; thus, the nominal height and diameterwere 12 and 6 mm, respectively.28 Sample rods were ob-tained with a Teflon® mold with cylindrical holes of 15 mmin height and 6 mm in diameter. Typically, five to six spec-imens were prepared from each batch and specimens withvisual flaws were discarded. The specimens were cut with aVC50 Varicut diamond saw (Leco Corp., Toronto, Canada)to produce 12-mm long cylinders with smooth, flat ends. Foreach cement set, between 4 and 6 specimens were tested.

Scanning Electron Microscopy. Scanning electron mi-croscopy (SEM) was performed with a JEOL 840A instru-ment, with an EDAX Phoenix system for image acquisition.Sample coating was performed using a Hummer VI sputter-coater with Au/Pd alloy target. SEM micrographs were re-corded using an acceleration voltage of 10 kV secondaryelectrons.

Micro Computed Tomography. Data were acquired on aSkyScan 1072 instrument (SkyScan, Aartselaar, Belgium),using cylindrical test specimens placed vertically in a preci-sion object manipulator, which was located between the X-Ray source, a microfocus sealed X-ray tube operating at avoltage of 100 kV and current of 98 �A, and the detector, anX-ray charge-coupled device camera. The specimen was ro-tated stepwise around its longitudinal axis, at a 0.9° angle,and images were acquired using a step exposure duration of2.240 s. A total of 206 images were recorded with 25�magnification at a spatial resolution of 10.95 �m. Crosssections along the length of the specimen were reconstructedusing Cone-Beam Reconstruction Software (SkyScan), with adistance between each cross section of 21.9 �m. The voxelsize of the images was 10.95 � 10.95 � 21.9 �m3. CTan andTview Software (both from SkyScan) were used to analyzethe sample.

Compression Testing. The uniaxial compression testswere performed on a servo-hydraulically-driven universal

materials testing machine (MTS Model 810, MTS Corp.,Minneapolis, MN), with a load frame rated for a maximumload of 25 kN, at a crosshead speed of 20 mm/min (ca. 2.78%strain/s), in ambient laboratory air (22 � 2°C). Treatment ofthe test results was as specified in ASTM F451–86.28

RESULTS AND DISCUSSION

Characterization of PMMA-graft-PHB Copolymers

From 1H NMR and GPC measurements, the graft copolymerswere found to contain 50 mol % of hydroxybutyrate (HB)units (47 wt % of PHB). The number-average molecularweight and polydispersity ratio, both measured by GPC, were63,400 and 1.5, respectively.

Morphology and Particle Size of the Polymer Powders

Figure 2 shows SEM micrographs of the polymer powdersused in the preparation of the bone cements, with PMMA-graft-PHB in Figures 2(A, B), and the powder from the AKZkit in Figures 2(C,D). The two polymer powders differed inboth particle morphology and size. The shape and size of thegraft copolymer particle were irregular, with an overall flakyappearance as seen in Figure 2(A). Some flocculation oraggregation was observed at higher magnification, giving theparticle surface a rough appearance [Figure 2(B)]. The par-ticle size was �100 �m. On the contrary, the powder fromAKZ was composed of polymer spheres of about 45 �m indiameter, partially covered with fluffy material [Figures2(C,D)]. From the micrographs in Figures 2(B, D), which aretypical for each sample, it can be concluded that the specificsurface area of the particles is higher for PMMA-graft-PHBthan for AKZ powder.

This difference in morphology can be explained by themethod employed to produce the particles. PMMA-graft-PHB particles were precipitated from solution, whereas theAKZ polymer powder was obtained from aqueous suspensionpolymerization,1 which allows the production of particles ofcontrolled sizes.29 The fluffy material around the AKZ poly-mer spheres could be the additives to the polymer: the ra-diopacifier (barium sulfate) and initiator (BPO) particles,1 orthe antibiotic. It could have also resulted from the suspensionpolymerization: the observed material could be stabilizers forthe particles.

Choice of Bone Cement Formulation

To evaluate the ability of the cement paste to be injected formolding, different formulations were prepared by varying theproportion of graft copolymer PMMA-graft-PHB in the pow-der component. BPO was also added to the powder, to keepthe same BPO to polymer ratio as for the AKZ formulation.The powder compositions can be found in Table I. Thepowder component has the capacity to swell and partiallydissolve in the liquid component, and therefore, the cementtexture and its ability to be processed were found to differ

8 NGUYEN AND MARCHESSAULT

with the polymer composition. For test T100, the mixing ofthe liquid monomer with the powder resulted in a materialthat resembled wet sand and remained very friable with time.In the case of test T50, the resulting material resembled a softsponge that could not be processed as bone cement. Whentest T10 was performed, a paste could be obtained aftermixing with the liquid monomer and hardened over time in asimilar manner to pure AKZ. For test T25, the mixing of thepowder with the liquid monomer produced a somewhat pow-dery paste, which hardened with time into a consistency thatresembled volcanic rock.

Two different formulations containing PMMA-graft-PHB,F10, and F20, described in Table I, were subsequently eval-uated by micro computed tomography (micro CT) to examinethe microarchitecture, and were also evaluated for their me-chanical properties using compression testing.

The differences in solubility of the polymer powders in theliquid monomer when T100, T50, T10, and T25 were usedwith one of the powder from AKZ may be attributed todifferences in the different chemical nature of the polymers inthe powders. In T100, T50, T10, and T25, the graft copoly-mers contained MMA and PHB segments with the HB unitsrepresenting 50 mol % of the comonomer molar ratio. TheAKZ powder contained the homopolymer PMMA mixedwith a copolymer of PMMA and polystyrene ; its styrenecontent was evaluated using 1H NMR spectroscopy, and wasfound to be 2.4 mol % repeat unit of total polymer material.

When PHB* was mixed with MMA, it was observed that themonomer swelled but did not dissolve PHB*, producing apaste at 2 g of polymer per mL of monomer, and a suspensionwith swollen particles at 0.1 g of polymer per mL of mono-mer. The polymers may also have different sorption abilitiesbecause of differences in their particle specific areas. In thecase of AKZ cement, the liquid monomer added during thecement preparation represented one third of the total mass ofthe material (manufacturer’s recommended ratio). WithPMMA-graft-PHB particles, this low amount of liquid mayhave sorbed on their rough surface. Therefore, it may nothave been in sufficient amount to dissolve the polymer par-ticles and to form a continuous phase allowing paste forma-tion and MMA polymerization around the partially dissolvedparticles. Abboud et al. reported a similar problem of cementthat could not be processed, for acrylic bone cement formu-lations containing silanated alumina particles.9,30 When theseparticles were in high concentration, or with high specificsurface area, it was observed that the quantity of liquidcomponent was insufficient to wet the alumina particle sur-face and to dissolve the PMMA beads, resulting in a cementlooking like wet sand.9,30

Porosity of the Bone Cements

Bone cements should contain a minimal number of pores ofa restricted size.2 Micro CT analysis provided microarchitec-

Figure 2. Scanning electron micrographs of the polymer particles used for bone cement preparation:(A and B) PMMA-graft-PHB, (C and D) powder from an Antibiotic Simplex® kit.

9COMPRESSION TESTING OF PHB ENRICHED ACRYLIC BONE CEMENTS

tural information on the localization of the pores as well astheir size, and the cement specific surface area in the differentformulations. The latter was defined as the ratio of the surfacearea to the total volume of the solid material. The results fortotal volumetric porosity, specific surface area, average poresize, and their distribution can be found in Table II. F10 andreference F0 presented similar total porosity and specificsurface area. In contrast, both values for these parameterswere higher for F20. The mean porosity of Simplex P ce-ments, the antibiotic-free version of AKZ, was reported at12.5%, when produced by manual mixing at 120 rpm with theliquid monomer previously cooled at 0°C.2 The mean poros-ity of the samples was therefore higher than the mean poros-ity of Simplex P cements prepared under similar conditions.The average pore sizes for the three formulations were low, inthe range of micropores, which are pores of diameter about0.1–1 mm. The discrepancy in pore size between F0, F20, andF10 may be due to experimental error.

Compression Testing

After micro CT analysis, the test specimens were subjected tocompression testing. The ASTM F451 “Standard Specificationfor Acrylic Bone Cement”28 was used as a guideline for thesemechanical experiments: the specimen dimensions, compressiontesting method, and the analyses following the ones indicated inthe standard. Standard aging time is 24 � 2 h, and defined as thetime elapsed between the start of cement mixing and the com-

pression tests.28 For the samples tested, the cement aging timewas longer—48 h, 42–46 h, and 40–44 h for F0, F10, and F20,respectively—because of the time needed to run micro CTanalyses on the samples before compression testing. In addition,the experiments, from the preparation of the bone cement to thecompression testing, were performed at room temperature[(20 � 2)°C for bone cement preparation and molding, (22 �2)°C for analyses] and room humidity, while the standard oper-ating temperature and humidity are ,respectively, (23 � 2)°Cand (50 � 10)%.28

Values of compressive strength, strain, and modulus, Ec,were calculated from the engineering stress–strain curves.2,28

According to ASTM F451 standard, the compressive strengthwas taken as the stress value first obtained,that is at the lowerstrain value, between the UCS, which was the maximumvalue of stress attained, and the yield strength at 2.0% off-set.28 For all the samples, the UCS values were found atlower strain values than the ones for the yield strengths, andwere therefore chosen as the compressive strength values.The different parameters are shown in Figure 3 for a sampleof formulation F10, with the strain corresponding to the UCSbeing the ultimate compressive strain, �maxc.

Results for the three formulations are shown in Table III.UCS values were found very close to one another, showinglittle influence of the PMMA-graft-PHB on the compressivestrength of the bone cement prepared here. In addition, theUCS of all the cements exceeded the minimum level stipu-lated in ASTM F-451, which is 70 MPa.28 The present resultswere within the range reported in the literature for AKZ,92 � 10 MPa, as determined in accordance with ISO 5833,1

which is another standard that is used in determining thecompressive properties of acrylic bone cements.2,9,30 Ec val-ues for F10 and F20 were quite similar, around 2400 MPa,and slightly (5%) higher than the one of reference F0, at 2270

Figure 3. Engineering stress–strain curve for a sample of formulationF10, when loaded in quasi-static compression.

TABLE II. Porosity of Bone Cements with Formulations F0, F10, and F20

Cement BatchNo.

VolumetricPorositya (%)

Specific SurfaceAreaa (mm�1)

Average PoreSize (mm)

Pore Size Distribution,Xvol (%)b

F0 13.0 � 3.0 10.8 � 2.0 0.15 78F10 13.5 � 3.0 11.0 � 2.0 0.30 76F20 16.9 � 3.0 13.8 � 2.0 0.13 83

a The results are given as mean � standard deviation.b Fraction of the total volume of pores, which are pores of diameter smaller than 0.20 mm.

TABLE III. Compressive Properties of Bone CementFormulations F0, F10, and F20

CementBatchNo.

UltimateCompressive

Strength (MPa)

UltimateCompressive

Strain(%)

CompressionModulus

(MPa)

F0 84.3 � 1.2 5.1 � 0.1 2270 � 180F10 84.5 � 7.0 5.2 � 0.3 2380 � 150F20 84.8 � 3.4 5.1 � 0.2 2390 � 70

The results are given as mean � standard deviation.

10 NGUYEN AND MARCHESSAULT

MPa. Therefore, bone cements containing up to 13.5 wt % ofPMMA-graft-PHB were slightly stiffer, or more resistant tochange in volume, than pure AKZ.

Compression properties of bone cements comprising de-gradable polymers have been reported, with performancesvarying with the composition. Espigares et al. have studiedthe behavior in compression of bone cements with a solid orpowder component constituted of corn starch/cellulose, hy-droxyapatite, and BPO, and a monomeric liquid of MMA,acrylic acid, and N-dimethylaminobenzyl alcohol.14 For ce-ments made from a solid to liquid weight ratio of 55 to 45, itwas shown that the UCS and Ec were similar or higher thanthose of commercial acrylic bone cements, ranging fromabout 57–98 MPa and 1330–2180 MPa, respectively, fordifferent fractions of hydroxyapatite.14 All the bone cementstested at this solid to liquid ratio displayed UCS superior tothe minimum level required by ASTM F-451, except onewith 20 wt % of nonsintered hydroxyapatite in the powdercomponent. Interestingly, the formulation that obtained thebest results contained 20 wt % of sintered hydroxyapatite inthe powder component.14 Compression testing was also re-ported for cements composed of a 30 wt % matrix ofcrosslinked PPF and a 70 wt % mixture of calcium carbonateand tricalcium phosphate.17 Different formulations using asthe polymer component (the rest being the calcium salt fillers)either divinyl-terminated, diol-terminated, or diepoxide-ter-minated PPF oligomers mixed with either MMA, or N-vinylpyrrolidone as crosslinking agents, or methyl pyrrolidone andethyl lactate as diluents were prepared and characterized.17 Itwas shown that only the UCS of a mixture of divinyl-terminated PPF oligomers with MMA and N-vinyl pyrroli-done as crosslinking agents met the ASTM F-451 require-ment, at a high value of 129.7 MPa.17 All the other compo-sitions presented UCS values lower than 70 MPa, rangingfrom 2.8 to 60.4 MPa.17

CONCLUSIONS

The powder of commercially available acrylic bone cement,AKZ, was partially replaced with PMMA-graft-PHB copol-ymer at different fractions of the cement. At higher contentthan 13.5 wt % of graft copolymers in the cement (corre-sponding to 20 wt % of copolymers in the powder compo-nent), the resulting cement could not be processed, probablybecause of the difference in solubility of the two polymers. Itcould also be due to the morphology of the graft copolymerparticles. The volumetric porosity of cements containing thegraft copolymer up to 13.5 wt % ranged from 13.5 to 16.9%,while their UCSs exceeded the minimum value as stipulatedin ASTM F451 (70 MPa) and were similar to that of AKZ.Work on determining other relevant properties, such as set-ting time, doughing time, maximum exotherm temperature,mechanical performance like tension, fatigue life, fracturetoughness, degradability, and biocompatility of cements con-taining this graft copolymer, is underway.

Dr. Gamal Baroud is specially thanked for helpful discussionsand the gift of the Antibiotic Simplex® AKZ kits, originally donatedby Stryker Canada LP, Howmedica Osteonics. We thank Dr. JanetHenderson and Dr. Thomas Steffen for their useful comments on themanuscript. Dr. Gregory Chauve is thanked for his help in SEM.Thanks are addressed to the McGill Centre for Bone and PeriodontalResearch, in particular Lydia Malynowski and Rujuan Huo for thetrimming of the bone cement test specimens, and Jean-SebastienBinette for performing the micro CT tests. Dr. Abdelbaset El-Wazriand Dr. Gregg Stewart are thanked for their assistance in thecompression testing. The technical help of Georges Kopp, AlfredKluck, and William Bastian is appreciated.

REFERENCES

1. Kuhn K-D. Bone Cements, Up-to-Date Comparison of Physicaland Chemical Properties of Commercial Materials. Berlin:Springer; 2000.

2. Lewis G. Properties of acrylic bone cement: State of the artreview. J Biomed Mater Res Part B: Appl Biomater 1997;38:155–182.

3. Charnley J. Anchorage of the femoral head prosthesis to theshaft of the femur. J Bone Joint Surg 1960;42B:28–30.

4. Lautenschlager EP, Stupp SI, Keller JC. Structure and proper-ties of acrylic bone cement. In: Ducheyne P, Hastings GW,editors. Functional Behavior of Orthopedic Biomaterials. BocaRaton: CRC Press; 1984. p 87–119.

5. Garfin SR, Yuan HA, Reiley MA. New technologies in spine:Kyphoplasty and vertebroplasty for the treatment of painfulosteoporotic compression fractures. Spine 2001;26:1511–1515.

6. Hardouin P, Fayada P, Leclet H, Chopin D. Kyphoplasty. JointBone Spine 2002;69:256–261.

7. Hardouin P, Grados F, Cotten A, Cortet B. Should percutaneousvertebroplasty be used to treat osteoporotic fractures? An up-date. Joint Bone Spine 2001;68:216–221.

8. Baroud G, Matsushita C, Samara M, Beckman L, Steffen T.Influence of oscillatory mixing on the injectability of threeacrylic and two calcium-phosphate bone cements for vertebro-plasty. J Biomed Mater Res Part B: Appl Biomater 2004;68B:105–111.

9. Abboud M, Vol S, Duguet E, Fontanille M. PMMA-basedcomposite materials with reactive ceramic fillers: Part III. Ra-diopacifying particle-reinforced bone cements. J Mater Sci:Mater Med 2000;11:295–300.

10. Hasenwinkel JM, Lautenschlager EP, Wixson RL, Gilbert JL. Anovel high-viscosity, two-solution acrylic bone cement: Effectof chemical composition on properties. J Biomed Mater Res1999;47:36–45.

11. Bunel C, Le Saux V, Vairon J-P. Cement for orthopedic pros-thesis, made of acrylic polymers. Fr. Patent 2,670,114, 1992 (toScience et Medecine, France).

12. Bunel C, Le Saux V, Vairon J-P. Biodegradable acrylic cementfor orthopedic prosthesis. Fr. Patent 2,670,115, 1992 (to Scienceet Medecine, France).

13. Jiang Y, Roby M. Biodegradable bone cement. U.S. Patent5,847,046, 1998 (to United States Surgical Corporation, USA).

14. Espigares I, Elvira C, Mano JF, Vazquez B, San Roman J, ReisRL. New partially degradable and bioactive acrylic bone ce-ments based on starch blends and ceramic fillers. Biomaterials2002;23:1883–1895.

15. Boesel LF, Mano JF, Reis RL. Optimization of the formulationand mechanical properties of starch based partially degradablebone cements. J Mater Sci: Mater Med 2004;15:73–83.

16. Boesel LF, Reis RL. Hydrophilic matrices to be used as bioac-tive and degradable bone cements. J Mater Sci: Mater Med2004;15:503–506.

17. Domb AJ, Manor N, Elmalak O. Biodegradable bone cementcompositions based on acrylate and epoxide terminated

11COMPRESSION TESTING OF PHB ENRICHED ACRYLIC BONE CEMENTS

poly(propylene fumarate) oligomers and calcium salt composi-tions. Biomaterials 1996;17:411–417.

18. Wise DL, Gresser JD, Trantolo DJ, Hsu Y-Y. Bioerodiblepolymeric semi-interpenetrating network alloys for surgicalplates and bone cements, and method for making same. U.S.Patent 6,071,982, 2000 (to Cambridge Scientific, Inc., Belmont,MA).

19. Brandl H, Bachofen R, Mayer J, Wintermantel E. Degradationand applications of polyhydroxyalkanoates. Can J Microbiol1995;41:143–153.

20. Doi Y. Microbial Polyesters. New York: VCH; 1990.21. Hocking PJ, Marchessault RH. Polyhydroxyalkanoates. In:

Kaplan DL, editor. Biopolymers from Renewable Resources.Berlin: Springer; 1998. p 220–248.

22. Zinn M, Witholt B, Egli T. Occurence, synthesis and medicalapplication of bacterial polyhydroxyalkanoate. Adv Drug De-livery Rev 2001;53:5–21.

23. Knowles JC. Development of a natural degradable polymer fororthopaedic use. J Med Eng Technol 1993;17:129–137.

24. Reusch RN. Low molecular weight complexed poly(3-hydroxy-butyrate): A dynamic and versatile molecule in vivo. Can JMicrobiol 1995;41:50–54.

25. Seebach D, Brunner A, Bachmann BM, Hoffmann T, KuhnleFNM, Lengweiler UD. Biopolymers and Oligomers of (R)-3-

hydroxyalkanoic Acids—Contributions of Synthetic OrganicChemists. Germany: Ernst Schering Research Foundation;1995. p 7–98.

26. Knowles JC, Hastings GW, Ohta H, Niwa S, Boeree N. Devel-opment of a degradable composite for orthopaedic use: In vivobiomechanical and histological evaluation of two bioactive de-gradable composites based on the polyhydroxybutyrate poly-mer. Biomaterials 1992;13:491–496.

27. Nguyen S, Marchessault RH. Synthesis and properties of graftcopolymers based on poly(3-hydroxybutyrate) macromono-mers. Macromol Biosci 2004;4:262–268.

28. American Society for Testing and Materials (ASTM). Specifi-cation F451–86. Standard Specification for Acrylic Bone Ce-ment, PA: ASTM; 1996. p 82–88.

29. Grulke EA. Suspension polymerization. In: Mark HF, BikalesNM, Overberger CG, Menges G, editors. Encyclopedia of Poly-mer Science and Engineering. New York: Wiley; 1989. p 443–473.

30. Abboud M, Casaubieilh L, Morvan F, Fontanille M, Duguet E.PMMA-based composite materials with reactive ceramic fillers:IV. Radiopacifying particles embedded in PMMA beads foracrylic bone cements. J Biomed Mater Res Part B: Appl Bio-mater 2000;53:728–736.

12 NGUYEN AND MARCHESSAULT