an in vitro optimized injectable calcium phosphate cement for augmenting screw fixation in...

An In Vitro Optimized Injectable Calcium Phosphate Cement forAugmenting Screw Fixation in Osteopenic Goats

Kwok Sui Leung, Wing Sum Siu, Siu Fai Li, Ling Qin, Wing Hoi Cheung, Kam Fai Tam, Pauline Po Yee Lui

Department of Orthopaedics and Traumatology, Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, NT,Hong Kong Special Administrative Region of the People’s Republic of China

Received 11 July 2005; revised 11 August 2005; accepted 15 August 2005Published online 16 November 2005 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jbm.b.30467

Abstract: This study reports the proportioning and standardized mixing procedures forpreparing a hydroxylapatite cement (tetracalcium phosphate and dicalcium phosphate) ofdesired viscosity and mechanical strength reproducibly for application in trauma surgery. Thebehavior and the biomechanical properties of the resulting bone cement in screw augmenta-tion were then evaluated in our osteopenic goat model. The use of a shaker standardized themixing procedure. The optimal volume of Na2HPO4 used to prepare the injectable cement was0.45 mL/g, with averaged in vitro compressive strength of 48.29 � 5.62 MPa. Histology showedincreasing tightly-coupled bone apposition on the cement surface without fibrous encapsula-tion as observed in the screw-only controls with time in the osteopenic goat model. The cementincreased the initial screw pull-out force (54.7%, p � 0.005) significantly and the energyrequired to failure (54.7%, p < 0.05) significantly, and remained higher than the screw-onlycontrols after 3 months (9.8% and 20.2%, respectively) and 6 months (20.2% and 44.7%,respectively). These results imply potential in the prevention of interfacial micromotions andsubsequent fibrous tissue formation at the implant–bone interface resulting in a decreasedrisk of implant failure. The optimized cement in this study may serve as a good candidate foraugmenting fixation of osteoporotic bone. © 2005 Wiley Periodicals, Inc. J Biomed Mater Res Part B:Appl Biomater 78B: 153–160, 2006

Keywords: calcium phosphate(s); animal model; fracture fixation; osteoporosis; tetracal-cium phosphate; dicalcium phosphate

INTRODUCTION

Bone cement is commonly used in orthopaedic surgery, forexample, augmentation of implant fixation in osteoporoticfractures, vertebroplasty, and total joint arthroplasty to in-crease the load-carrying capacity of the prosthesis–bone sys-tem.1–6 It is also frequently used in craniofacial reconstruc-tion to recontour the defect.7,8 Recent advances in bonecement technology result in new formulation that is biocom-patible, osteoconductive, nonexothermic during cement set-ting, and will not induce chemical necrosis and inflammatoryreactions.6 Importantly, it will also gradually resorb withbone growth.6

The quality of cement produced depends very much on thepreparation, mixing, and delivery techniques.6,9–16 Completemixing is very important. This is especially the case forcalcium phosphate cement, which is recognized to be moredifficult to mix and inject than acrylic cement. During mixing

and transferal from the mixing chamber to the syringe or gun,air bubbles can be trapped in the cement paste. This results inthe appearance of macropores (with diameters in excess of 1mm) and micropores (with diameters between 0.1 and 1 mm)in the paste after hardening. Cement porosity has been im-plicated in the early loosening of cemented total hip arthro-plasties.17–19 The wide variation in cement quality due toinsufficient mixing and air entrapment significantly affectsthe final composition and the usefulness of bone cement,especially in areas requiring the bone cement to providemechanical support.9,10 Studies to achieve complete mixingand reduce entrapment of air in cement using different mixingtechniques are current active research areas.9–15 Besides,bone cement sets too rapidly, which does not allow sufficienttime for preparation and application. Inconvenience in prep-aration and frequent transfer of bone cement paste duringpreparation, mixing, and delivery to surgical site both in-crease the chance of bacterial contamination and infection.

Injectable hydroxylapatite bone cement is primarily foruse in the recontouring of nonweight bearing craniofacialskeletal defects.7,20 There have been fewer reports of its usein augmenting implant fixation or in weight-bearing re-gion.4,21–24 A given property of bone cement is both formu-

Correspondence to: K. S. Leung (e-mail: [email protected])Contract grant sponsor: Chinese University of Hong Kong; contract grant number:

2004.1.077

© 2005 Wiley Periodicals, Inc.

153

lation-specific and mixing method-specific.15 To facilitate theexploration of bone cement’s application in augmenting im-plant fixation, standardized procedures for its preparation isvery important, especially in fixation enhancement of osteo-porotic fractures, the incidences of which increase dramati-cally with aging.25 This study reports the standardized pro-cedures and proportioning for preparing a hydroxylapatite(HA) bone cement, which is formed by the dissolution andprecipitation of tetracalcium phosphate and dicalcium phos-phate in an aqueous system under isothermic reaction and atphysiological pH [CaHPO4 � Ca4(PO4)2O 3 Ca5(PO4)3OH],for application in trauma surgery. The behavior and thebiomechanical properties of the resulting bone cement inscrew augmentation were then evaluated using our estab-lished osteopenic goat model.26

MATERIALS AND METHODS

Determination of the Mixing Ratio of Bone CementPowder and Na2HPO4

Five grams of bone cement powder (a mixture of tetracalciumphosphate and dicalcium phosphate anhydrous; Ca/P ratio:1.65–1.67; from Stryker Leibinger GmbH & Co. Kg,Freiburg, Germany) was mixed with 1.5, 1.75, 2.0, 2.25, or2.5 mL of Na2HPO4 (0.25M) at room temperature (22°C). Asterile 20-mL syringe was preweighed. Afterwards, the bonecement powder and Na2HPO4 were added to the syringe andmixed inside a shaker (Stryker Leibinger GmbH & Co. Kg,Freiburg, Germany) at 1000 rpm for 13 s. The total weight ofthe syringe and the mixed bone cement was recorded. Thebone cement was then allowed to set for different periods oftime: 0, 5, 10, and 20 min after mixing. At 0, 5, 10, and 20min, the extrudability of the mixed bone cement preparedwith different volume of Na2HPO4 was tested.27

Indentation Test

Mechanical property of the bone cement prepared with theoptimal powder to Na2HPO4 ratio was evaluated by indenta-tion test.28 This was used as the end point of optimizationbefore in vivo evaluation. After mixing, the bone cement wascured (at 37°C for 24 h, at 100% humidity) to produce a 10mm (diameter) � 10 mm (height) cement block. A 2.5-mmdiameter indentation rod connected to a mechanical testingmachine (H25K-S, Hounsfield Test Equipment, Salfords,Redhill, UK) was loaded on the center of the cement block,at 10 mm/min. The failure stress was recorded for six sam-ples.

Osteopenic Goat Model

This part of the study was approved by the Animal EthicsCommittee of the Chinese University of Hong Kong (04/003/ERG). Nineteen Chinese mountain goats were used, withskeletal maturity radiologically confirmed by closure of

growth plate at the proximal tibia. The osteopenic animalmodel was established by ovariectomy plus low calcium dietaccording to our protocol developed previously.26 Briefly,under halothane anaesthesia, a window was made in themesovarium caudal to the ovarium vessels. The ovarian pedi-cle was triple clamped and severed. The clamp most distantfrom the ovary was then removed so that the pedicle ligaturecould be passed into its groove. Nonabsorbable suture mate-rial was used for the ligature and the ovary was excised.Similar operation was performed for the other ovary. Finally,the wound was closed by suture. All the ovariectomized goatswere fed with low calcium diet (50% of food pellet, GlenForrest Stockfeeders, Australia; plus 50% chaff, WheatenChaff, O’Driscoll, Australia). The total calcium content of thediet was 0.5% compared to 1.1% in normal diet.

Both the femoral condyles and the lumbar vertebraeL2–L5 were selected for studying the effect of bone cementon screw fixation. The healing and remodeling of bonearound the screw threads and bone cement were evaluatedhistologically at the femoral condyles, while the mechanicalstrength of the screw augmented with bone cement wasassessed by screw pull-out tests.

Screw Fixation and Augmentation by Bone Cement

There were two testing groups. Two randomly selected lum-bar vertebrae of each goat served as treatment groups withboth screw insertion and bone cement injection while theother two lumbar vertebrae of the same goat acted as controlswith screw insertion only. Under general anesthesia, thelumbar vertebrae were exposed with left peritoneal approach.A 2.7 mm (diameter) � 12 mm (depth) hole was then drilledinto the vertebral body of each lumbar vertebra, using a2.7-mm drill bit. The drilled hole was irrigated with salineand dried with gauze. Meanwhile, injectable bone cementwas prepared according to the optimal procedures determinedearlier. For the experimental group, 0.15 mL of bone cementwas applied into the drilled hole. A stainless stain–titaniumcancellous screw of 20 mm in length and 4 mm in threaddiameter was inserted into the hole afterwards, with sixcycles of rotation before the cement hardened. For the controlgroup, only the cancellous screw was inserted. Soft tissueswere closed by suture in layers. Similar procedures wereperformed at the femoral condyles. A longitudinal incisionwas made on the lateral side of the femur. After exposing thefemoral condyle, a 2.7-mm hole was drilled unicortically andperpendicularly to the long axis of the bone with a 2.7-mmdrill bit. At 1 week (baseline), 3 months, and 6 months afterscrew fixation, the operated goats were euthanized by over-dose intravenous administration of pentobarbital (60 mg/kg).The lumbar vertebrae were harvested for screw pullout testand the femoral condyles were excised for histological anal-ysis, respectively, after removal of the excess soft tissue.

Screw Pull-Out Test

Screw pull-out force and the associated energy required tofailure were determined using the above-mentioned material

Journal of Biomedical Materials Research Part B: Applied BiomaterialsDOI 10.1002/jbmb

154 LEUNG ET AL.

testing machine. The bone specimen was embedded in aperformance polymer (UREOL 2020, Ciba, Hong Kong)exposing only the screw head. A custom-made jig was usedto grasp the screw head and the specimen was held rigidly atthe other end of the jig. Screw inside the specimen was pulleduntil failure, at a constant speed of 10 mm/min. Force anddisplacement were obtained in real time, using a computer-ized data acquisition system with sampling rate of 10 Hz. Thepull-out force (N) was determined as the maximum point onthe force–displacement curve where failure occurred. Theenergy required to failure (J) was calculated based on the areaunder the curve.28 Results were average of 8–10 specimensfor each time point.

Histological Analysis

The specimens, with the screws left in situ, were dehydratedsequentially with 70% ethanol for 3 days, 90% ethanol for 2days, 100% ethanol for 3 days, 100% ethanol and xylene(1:1) for 2 days, and finally xylene twice for 2 days. Thespecimens were then infiltrated with MMA, according to ourwell-established protocol.29 The embedded specimens weresectioned in a longitudinal fashion parallel to the axis of theimplant to a thickness of 500 �m, using a saw microtome(Leica SP1600, Leica Instruments, Nussloch, Germany). Thesections were further grinded and polished to a thickness of200 �m, using a grinder/polisher (RotoPol-21, Struers, Den-mark). The sections were processed for routine histology andexamined with a light microscope (Leica DMRXA2, LeicaMicrosystems Wetzlar GmbH, Germany) and polarized lightmicroscope (Leica DMRB, Leica Microsystems WetzlarGmbH, Germany).

Statistics

The results of both the indentation test and pull-out test wereexpressed as means � SD. Nonparametric Wilcoxon’ssigned-ranks test was used to compare the pull-out force andenergy between the experimental and control groups. Non-parametric Kruskal–Wallis test was used to compare thedifference between different time points of each group. Allthe data analysis was done using SPSS version 11.0 (SPSS

Inc., Chicago, IL, USA). The level of significance was set atp � 0.05 (2-tailed).

RESULTS

Optimization of Bone Cement

Table I shows the results for the extrudability of the bonecement. Na2HPO4 at 0.3 mL/g was insufficient to prepare theinjectable bone cement. No cement could be extruded fromthe 20-mL syringe. The homogeneity was so poor and somedry powder remained after mixing. When the bone cementpowder was mixed with Na2HPO4 at 0.35 and 0.4 mL/g,similar homogeneities were observed. Although the later ex-hibited a better extrudability (95.2%) than that of the former(91.8%) immediately after mixing, both of them could not beextruded from the syringe smoothly after 5 min. Less than ahalf of the bone cement could be squeezed out (49.2% and39.7% for 0.35 and 0.4 mL/g of Na2HPO4, respectively) at 5min. Under these two mixing protocols, no bone cementcould be forced out after 10 min of setting. Good viscosityand extrudability of injectable bone cement were obtainedwhen Na2HPO4 was mixed with bone cement powder at 0.45mL/g. Nearly 97% of bone cement could be extruded imme-diately after mixing while 64.6% was available for use 10min after mixing. When Na2HPO4 was mixed with bonecement powder at 0.5 mL/g, the mixed bone cement becametoo liquid for application. For the indentation test, the aver-aged failure stress of the injectable bone cement prepared inthe optimal volume of Na2HPO4 (0.45 mL/g) was 48.29 �5.62 MPa (ranging between 41.61 and 56.31 MPa).

Handling of Bone Cement In Vivo

On the basis of the optimal in vitro conditions, as describedearlier, there was sufficient time for handling the bone cementfor screw fixation in our animal model. Setting was notinfluenced by the presence of blood, although we dried thesurgical site briefly with gauze after drilling. No inflamma-tion or adverse tissue reaction up to 6 months after screwfixation and cement augmentation was observed in the ani-mals.

TABLE I. Extrudability of Injectable Bone Cement Prepared in Different Volume of Sodium Phosphate Solution (Na2HPO4)

Incubation time(minutes)

Weight of bone cement extruded (g)

1.5 (0.30 mL/g)a,b 1.75 (0.35 mL/g) 2.0 (0.40 mL/g) 2.25 (0.45 mL/g) 2.5 (0.50 mL/g)

0 0 (0)c 5.6 (91.8) 6.0 (95.2) 6.3 (96.9) 6.7 (98.5)5 3.0 (49.2) 2.5 (39.7) 6.0 (92.3) 6.7 (98.5)

10 0 (0) 0 (0) 4.2 (64.6) 6.6 (97.1)20 0 (0) 3.2 (47.1)

Data were expressed as weight (g) of bone cement extruded (%).a Volume (mL) of 5M Na2HPO4 for 5 g of bone cement (mL/g)b The initial net weight of bone cement (g) for 1.5, 1.75, 2.0, 2.25, and 2.5 mL of 5M Na2HPO4 are 5.7, 6.1, 6.3, 6.5, and 6.8 g, respectively.c Values in parentheses are expressed in percentage.


155INJECTABLE CEMENT FOR FIXATION OF OSTEOPENIC BONE

Descriptive Histology

At baseline (week 1), the bone cement filled most of the voidspace surrounding the implant, with much less void space inthe area surrounding the screw threads in the cement groupcompared with that in the control group [Figure 1(a,e)]. Nofibrous tissue was observed in the screw threads filled withbone cement except at the top of the screw head, which couldnot be filled with bone cement [Figure 1(e–f)]. There was nofragmentation of bone cement and it appeared homogeneous[Figure 1(d)]. About two-thirds of the surface of the bonecement was covered with a thin layer of bone at baseline(week 1). A tight contact existed between the bone and thecement surface [Figure 1(g)]. No intervening fibrous tissuelayer was present. The trabeculae were thin. At month 3,more bone was observed and most, if not all, of bone cementsurrounding the screw threads was covered with bone [Figure1(e)]. The woven bone showed a better alignment comparedwith that at week 1, which was dominated by broken bonefragments [Figure 2(e)]. The cement mantle remained intact.In some areas, penetration of bone into the periphery aroundoccasionally fragmented pieces of cement was observed [Fig-ure 1(h)]. At month 6, more bone was observed and theyspread along the bone cement surface [Figure 1(f)]. Theywere properly aligned and maturation started [Figure 2(f)]. Insome screw threads, the woven bone completely filled thespace with no intervening fibrous tissue [Figure 1(f)].

Different results were observed for the screw-only con-trols. At baseline (week 1), vast amount of empty space and

some broken bone fragments due to the drilling procedurewere seen in the bone bed–screw interface [Figure 1(a),arrows]. Fibrous tissue began to infiltrate the screw and boneinterface. At month 3, the screw threads were completelycovered with fibrous tissue with some voids between thefibrous tissue–implant interface [Figure 1(b), arrow]. Morebone was observed compared with baseline (week 1), butthey did not reach and contact with the screw [Figure 1(b)].At month 6, the screw threads were still covered with fibroustissue. Additional bone was observed and grew towards thescrew but it was hindered by the fibrous layer [Figure 1(c)].Despite this, better alignment of woven bone was observed after6 months compared to that after 3 months [Figure 2(b,c)].

Screw Pull-Out Test

The screw pull-out force of the bone cement group wassignificantly higher than the control, at all time points exceptat 3 months after implantation, which was marginally insig-nificant. There was an immediate 54.7% gain in screw pull-out force at baseline (week 1) after cement injection com-pared with the control group (Figure 3). However, the differ-ences between the bone cement and control groups were lessat 3 and 6 months after implantation (3 months: 9.8%; 6months: 20.5%) (Figure 3). Similar to pull-out force, theenergy required to failure in the bone cement group was alsosignificantly higher than that in the control group, at all timepoints [54.7%, 20.2%, and 44.7% for baseline (week 1), 3,and 6 months, respectively], indicating that there was an

Figure 1. Photomicrographs showing the trabecular bone around the implant without (a–c) and with(d–i) bone cement at different times after implantation with a light microscope. (a–f) �16. (g–i) �50. F,fibrous tissue; T, trabeculae; C, cement; arrows, space; S, screw.


156 LEUNG ET AL.

improvement in screw holding power immediately after bonecement injection (Figure 4). Both screw pull-out force andenergy increased with time in both groups and they weresignificant only for the control group (Figures 3 and 4). Theinsignificant increase in screw pull-out force and energy atmonth 3 and month 6 compared with baseline (week 1) wasdue to the initial gain in energy and pull-out force in thecement group at baseline (week 1), which reduced the differ-

ences when compared to the subsequent time points. Obser-vations during mechanical testing showed failure occurred inthe bone layer in both groups at all time points with morebone attached to the implant of the cement group.

DISCUSSION

This was the first study to investigate the long-term effect ofcalcium phosphate cement in osteoporotic fracture fixation,

Figure 2. Photomicrographs showing trabecular bone alignment around the implant without (a–c) andwith (d–f) bone cement at different times after implantation with a polarized light microscope (�16). F,fibrous tissue; T, trabeculae; C, cement; S, screw.

Figure 3. Screw pull-out force in the experimental (cement) andcontrol groups at different times after implantation.

Figure 4. Pull-out energy in the experimental (cement) and controlgroups at different times after implantation.



using an osteopenic goat model. Most of the previous studieson this formulation of tetracalcium phosphate–dicalciumphosphate bone cement were focused on its immediate bio-mechanical properties in screw augmentation in vitro4,23,24

and there had been few studies on its long-term effects invivo.21,22 Our results showed that the bone cement was highlybiocompatible and osteoconductive. There was a gain in thescrew pull-out force and energy immediately after cementfixation.

Before the application of bone cement in vivo, we haveoptimized the HA bone cement preparative procedures. Bonecement mixed at different powder-to-liquid ratios have dif-ferent degrees of porosity, strength, and load-bearing capac-ity.6,9,30 Our results showed that the optimal volume ofNa2HPO4 used to prepare the injectable bone cement was0.45 mL/g. There was a critical powder-to-liquid ratio, be-yond which the viscosity of the mixed bone cement decreaseddramatically. It was demonstrated that 0.25 mL of Na2HPO4

to 2.5 mL (i.e. 0.5 mL/g) resulted in a cement “paste” thatwas too watery for application. Accurate measurement ofliquid volume and cement powder was possible and importantfor success. In fact, this was different from the liquid-to-powder ratio suggested by the manufacturer’s instruction andliterature.20 Mixing the cement at 0.5 mL/g by hand wasusually recommended, which was found to be too liquid forapplication in our experiment. In Friedman et al.’s (1998)study, 0.25 mL/g of Na2HPO4 was suggested for hand-mix-ing the bone cement for craniofacial surgery, which was notpossible in our case as even 0.3 mL/g was not enough to mixthoroughly with the bone cement powder.20 Also, we used ashaker for mixing. Spillage of cement, compositional varia-tions, and poorly controllable setting times, which mightoccur because of hand mixing, thus could be avoided. Therewere many studies on the effects of different mixing systemsbased on different principles on the quality of bone ce-ment,9–15 and most of them showed improvement when com-pared to mixing by hand. Injection and manipulation of bonecement were possible up to 5 min after the cement started toharden and could not be extruded from the syringe com-pletely. This setting time was adequate for cement deliveryeven for complicated surgical procedures such as fracturefixation. Overall, the components were easy to handle, themixing procedure was simple, the viscosity of the cement wasconsistent and optimal to avoid extravasations. One limitationwas that the radio-opacity of the biomaterial was not goodenough for monitoring its postoperative distribution. Thismight be improved in the future study by using contrast-enhancing materials to increase the radio-opacity of calciumphosphate cement. The use of BaSO4, ZrO2, and some newestagents like iopromide to enhance the radio-opacity of calciumphosphate cement has been reported.31,32

The compressive strength of bone cement prepared in thisoptimal powder-to-liquid ratio was in the range of 41.61–56.31 MPa. This was similar to the results reported previ-ously, in which the 24-h wet compressive stress of the HAbone cement from the same manufacturer was reported to be51.0 � 4.5 MPa.33 Our previous study in goats showed that

the compressive strength of cancellous bones for normal andosteopenic goats were 42.01 � 1.93 MPa and 20.46 � 2.16MPa, respectively.26 Jansen et al.6 also mentioned that thecompressive strength of calcium phosphate cement had to begreater than 30 MPa for use in the human trabecular bone.The HA bone cement prepared according to our standardizedmixing ratio and procedures was thus suitable for augmenta-tion of implant fixation in osteoporotic trabecular bone, andcould be modified and evaluated in our osteopenic goatmodel.

The osteopenic goat model was established by ovariec-tomy plus low calcium diet.26 Our previous data showed thatsuch treatment could induce significant osteopenia at the iliaccrest, L2, L7, calcaneus and humeral head, deterioration oftrabecular microarchitecture, and mechanical properties ofthe cancellous bone in goats.26,34 The median screw pull-outstrength in both cortical and cancellous bones of the ovari-ectomized goats was lower than that in sham-operated goats,although the difference was significant only at the cancellousbones.35 The biological interaction of the osteopenic bonewith the screw was thus expected to be poorer than that innormal bone. We do not have data on the screw pull-outstrength augmented with bone cement in normal goat. How-ever, we expect it to be higher than that in osteopenic bone,based on the decrease in bone density and bone architecturalparameters in osteopenic bone observed in our previousstudy.26,34

In our experiments, the calcium phosphate bone cementdid not evoke any inflammatory reaction up to 6 months. Itwas found to be osteoconductive, with trabecular bone de-posited on the cement surface. No encapsulation by fibroustissue occurred between the bone cement and implant. Incontrast, fibrous tissue surrounded the implant for the controlgroup as there was no evidence of implant–bone contact at alltime points despite better alignment and hence maturation ofwoven bone at 6 months. These findings were consistent withseveral other studies working on calcium phosphate ce-ment.36–38 Micromovements of implant after implantationhave been suggested to be responsible for the formation offibrous tissue in the interface. This, compounded with thepoor architecture of osteoporotic bone, led to low pull-outforce and early implant failure. On the other hand, new boneanchored the hydroxyapatite–implant system in our experi-ment, and hence, the implant had higher pull-out force andbetter fixation under the same condition.39

Similar to many ex vivo experimental studies with differ-ent formulations of calcium phosphate cements for augmen-tation of screw fixation,24,40,41 the calcium phosphate cementused in the present study provided significant reinforcementof screw fixation immediately after surgery. A higher me-chanical strength immediately after operation was shown tobe able to maintain defect stability at the initial stage, whenlittle bone had grown to the defect site. Moreover, the me-chanical strength remained higher than the control group, atthe end of the study. The reduction in the difference in thepull-out force between the experimental and control groups at3 and 6 months found in the present study might be explained


158 LEUNG ET AL.

by new bone formation in the control group. The point offailure for the screw pull-out test occurred in the bone layer,indicating that the weak link did not lie in the screw–cementand cement–bone interfaces, which has been suggested as thebiomechanical weak links in the implant–bone and cement–bone system, by others in in vitro mechanical tests of bonecement.17,18 This might be explained by the high osteocon-ductivity of calcium phosphate cement prepared in the currentstudy and the reduction in porosity at the screw–cementinterface with in situ setting nature of bone cement.

The bone cement looked dense and homogeneous after 6months in those areas where bone cement could be seen. Thedense cement precluded bone ingrowth, as demonstrated inour study, and limited its clinical application to bone voidfiller.42 This was also supported by the findings of biome-chanical study in which both the energy required to failureand pull-out force leveled off after 3 months in both groups.This might be due to the nearly “complete” bone formation atmonth 3, while bone maturation occurred at month 6. Fibroustissue prevented further ingrowth of bone to the implant in thecontrol group, while it took years to completely replace thebone cement by bone in the cement group. Resorption ofbone cement to create macropores would help bone in-growth.43,44 This is an important material property because itpermits new bone tissue to grow into the defect. Cementformulations with resorbable crystals such as mannitol, sugar,hydroxypropyl-methylcellulose, or calcium sulfate dehydratethat create controllable porosity in the matrix in vivo and withgood enough mechanical properties at the early implantationstage were being investigated.23,24,42,44–46

While calcium phosphate cements can withstand compres-sive load, they have relatively low fracture strength, brittle-ness, and susceptible to fatigue failure. These might be thereasons for the limited use of calcium phosphate cement incomplex loading situations like screw augmentation. A sim-ple in vitro pull-out test mimics the clinical situation onlypartially. A dynamic fatigue test would be more suitable andshould be conducted in the future. The resorption character-istics of the bone cement was not investigated in the presentstudy. This information is important for further improvementof the mechanical as well as the biological properties of thebone cement for different individuals and applications indifferent anatomical regions. In patients with poor bone qual-ity or poor bone remodeling, it might not be necessary tocompletely replace the bone cement by bone as the newlyformed bone might not have the necessary mechanical prop-erties to hold the screw in place. The resorption characteris-tics of the bone cement will be evaluated in the future.

CONCLUSION

In conclusion, we reported the long-term biological behav-iour of HA bone cement in vivo after in vitro standardization.We have optimized the procedures to produce a consistentcement paste of desired viscosity, which set in a workabletime and provided good mechanical strength for use in

weight-bearing osteopenic regions. The calcium phosphatecement prepared according this protocol showed high osteo-conductivity and increased the screw pull-out force and en-ergy required to failure, when used in screw augmentation. Itthus has potential application in augmenting fixation of os-teoporotic bone.

REFERENCES

1. Elder S, FrankenBurg E, Goulet J, Yetkinler D, Poser R, Gold-stein S. Biomechanical evaluation of calcium phosphate ce-ment-augmented fixation of unstable intertrochanteric fractures.J Orthop Trauma 2000;14:386–393.

2. Font-Rodriguez DE, Insall JN, Scuderi GR. Survivorship ofcemented total knee arthroplasty. In Proceedings of the 64thAnnual Meeting, American Academy of Orthopeadic Surgeons,San Francisco, CA, 1997. p 151.

3. Simpson D, Keating JF. Outcome of tibial plateau fracturesmanaged with calcium phosphate cement. Injury 2004;35:913–918.

4. Belkoff SM, Mathis JM, Jasper LE, Deramond H. An ex vivobiomechanical evaluation of a hydroxyapatite cement for usewith vertebroplasty. Spine 2001;26:1542–1546.

5. Heini PF, Berlemann U, Kaufmann M, Lippuner K, FankhauserC, van Landuyt P. Augmentation of mechanical properties inosteoporotic vertebral bones—A biomechanical investigation ofvertebroplasty efficacy with different bone cements. Eur SpineJ 2001;10:164–171.

6. Jansen J, Ooms E, Verdonschot N, Wolke J. Injectable calciumphosphate cement for bone repair and implant fixation. OrthopClin North Am 2005;36:89–95.

7. Reddi SP, Stevens MR, Kline SN, Villanueva P. Hydroxyapatitecement in craniofacial trauma surgery: Indications and earlyexperience. J Craniomaxillofac Trauma 1999;5:7–12.

8. Friedman CD, Costantino PD, Synderman CH, Chow LC,Takagi S. Reconstruction of the frontal sinus and frontofacialskeleton with hydroxyapatite cement. Arch Facial Plast Surg2000;2:124–129.

9. 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 B 2004;68:105–111.

10. Burke DW, Gates EI, Harris WH. Centrifugation as a method ofimproving tensile and fatigue properties of acrylic bone cement.J Bone Joint Surg Am 1984;66:1265–1273.

11. Davies JP, Jasty M, O’Connor DO, Burke DW, Harrigan TP,Harris WH. The effect of centrifugating bone cement. J BoneJoint Surg Br 1989;71:39–42.

12. Topoleski LD, Ducheyne P, Cuckler JM. The effects of centrif-ugation and titanium fiber reinforcement on fatigue failuremechanisms in poly(methyl methacrylate) bone cement.J Biomed Mater Res 1995;29:299–307.

13. Norman TL, Kish V, Blaha JD, Gruen TA, Hustosky K. Creepcharacteristics of hand- and vacuum-mixed acrylic bone cementat elevated stress level. J Biomed Mater Res 1995;29:495–501.

14. Fritsch EW. Static and fatigue properties of two new low-viscosity PMMA bone cements improved by vacuum mixing.J Biomed Mater Res 1996;31:451–456.

15. Lewis G. Effect of mixing method and storage temperature ofcement constituents on the fatigue and porosity of acrylic bonecement. J Biomed Mater Res 1999;48:143–149.

16. Dunne NJ, Orr JF. Curing characteristics of acrylic bone ce-ment. J Mater Sci Mater Med 2002;13:17–22.

17. Davies JP, Kawate K, Harris WH. Effect of interfacial porosityon the torsional strength of the cement-metal interface. TransOrthop Res Soc 1995;20:713.



18. James SP, Jasty M, Davies J, Piehler H, Harris WH. A fracto-graphic investigation of PMMA bone cement focusing on therelationship between porosity reduction and increased fatiguelife. J Biomed Mater Res 1992;26:651–662.

19. Jasty M, Maloney WJ, Bragdon CR, O’Connor DO, Haire T,Harris WH. The initiation of failure in cemented femoral com-ponents of hip arthroplasties. J Bone Joint Surg Br 1991;73:551–558.

20. Friedman CD, Costantino PD, Takagi S, Chow LC. BoneSourcehydroxyapatite cement: A novel biomaterial for craniofacialskeletal tissue engineering and reconstruction. J Biomed MaterRes 1998;43:428–432.

21. Dickson KF, Friedman J, Buchholz JG, Flandry FD. The use ofBoneSource hydroxyapatite cement for traumatic metaphysealbone void filling. J Trauma 2002;53:1103–1108.

22. Blattert TR, Delling G, Weckbach A. Evaluation of an inject-able calcium phosphate cement as an autograft substitute fortranspedicular lumbar interbody fusion: A controlled, prospec-tive study in the sheep model. Eur Spine J 2003;12:216–223.

23. Lim TH, Brebach GT, Renner SM, Kim WJ, Kim JG, Lee RE,Andersson GBJ, An HS. Biomechanical evaluation of an inject-able calcium phosphate cement for vertebroplasty. Spine 2002;27:1297–1302.

24. Renner SM, Lim TH, Kim WJ, Katolik L, An HS, AnderssonGBJ. Augmentation of pedicle screw fixation strength using aninjectable calcium phosphate cement as a function of injectiontiming and method. Spine 2004;29:E212–E216.

25. Buckwalter JA, Heckman JD, Petrie DP. Aging of the NorthAmerican Population: New challenges for orthopaedics. J BoneJoint Surg Am 2003;85:748–758.

26. Leung KS, Siu WS, Cheung NM, Lui PY, Chow DH, James A,Qin L. Goats as an osteopenic animal model. J Bone Miner Res2001;16:2348–2355.

27. Khairoun I, Boltong MG, Driessens FCM, Planell JA. Somefactors controlling the injectability of calcium phosphate bonecements. J Mater Sci Mater Med 1998;9:425–428.

28. Qin L, Zhang M. Mechanical testing for bone specimens. In:Deng HW, Liu YZ, editors. Current Topics of Bone Biology.Singapore: World Scientific; 2005. p 177–212.

29. Qin L, Mak AT, Cheng CW, Hung LK, Chan KM. Histomor-phological study on pattern of fluid movement in cortical bonein goats. Anat Rec 1999;255:380–387.

30. Belkoff SM, Sanders JC, Jasper LE. The effect of the monomer-to-powder ratio on the material properties of acrylic bone ce-ment. J Biomed Mater Res 2002;63:396–399.

31. Stallmann HP, Faber C, Plokker HM, Wuisman PI. Biodegrad-able X-ray markers of controlled radio-opacity. Temporary po-sition measurements in bone. Acta Orthop Scand 2005;76:122–127.

32. Ginebra MP, Aparicio C, Albuixech L, Fernandez-Barragan E,Gil FJ, Planell JA, Morejon L, Vazquez B, San Roman J.

Improvement of the mechanical properties of acrylic bone ce-ments by substitution of the radio-opaque agent. J Mater SciMater Med 1999;10:733–737.

33. Chow LC, Takagi S, Constantino PD, Friedman CD. Self-setting calcium phosphate cements. Mater Res Symp Proc 1991;179:3–24.

34. Siu WS. An attempt to establish an osteoporotic animal model.MPhil Thesis, The Chinese University of Hong Kong, HongKong, 2000.

35. Siu WS, Qin L, Cheung WH, Leung KS. A study of trabecularbones in ovariectomized goats with micro-computed tomogra-phy and peripheral quantitative computed tomography. Bone2004;35:21–26.

36. Frankenburg EP, Goldstein SA, Bauer TW, Harris SA, PoserRD. Biomechanical and histological evaluation of a calciumphosphate cement. J Bone Joint Surg Am 1998;80:1112–1124.

37. Yuan H, Li Y, de Bruijn JD, de Groot K, Zhang X. Tissueresponses of calcium phosphate cement: A study in dogs. Bio-materials 2000;2:1283–1290.

38. Ooms EM, Wolke JGC, van der Waerden JPCM, Jansen JA.Use of injectable calcium-phosphate cement for the fixation oftitanium implants: An experimental study in goats. J BiomedMater Res B 2003;66:447–456.

39. Soballe K, Hansen ES, Brockstedt-Rasmussen H, Bunger C.Hydroxyapatite coating converts fibrous tissue to bone aroundloaded implants. J Bone Joint Surg Br 1993;75:270–278.

40. Lotz JC, Hu SS, Chiu DF, Yu M, Colliou O, Poser RD.Carbonated apatite cement augmentation of pedicle screw fix-ation in the lumbar spine. Spine 1997;22:2716–2723.

41. Moore DC, Frankenburg EP, Goulet JA, Goldstein SA. Hipscrew augmentation with an in situ-setting calcium phosphatecement: An in vitro biomechanical analysis. J Orthop Trauma1997;11:577–583.

42. Nilsson M, Fernandez E, Sarda S, Lidgren L, Planell JA. Char-acterization of a novel calcium phosphate/sulphate bone ce-ment. J Biomed Mater Res 2002;61:600–607.

43. Tamai N, Myoui A, Tomita T, Nakase T, Tanaka J, Ochi T,Yoshikawa H. Novel hydroxyapatite ceramics with an intercon-nective porous structure exhibit superior osteoconduction invivo. J Biomed Mater Res 2002;59:110–117.

44. Xu HHK, Takagi S, Quinn JB, Chow LC. Fast-setting calciumphosphate scaffolds with tailored macropore formation rates forbone regeneration. J Biomed Mater Res A 2004;68:725–734.

45. Markovic M, Takagi S, Chow LC. Formation of macropores incalcium phosphate cements through the use of mannitol crys-tals. Giannini S, Moroni A, editors. Bioceramics, Vol. 13.Bologna, Italy: Trans Tech; 2000. p 773–776.

46. Takagi S, Chow LC. Formation of macropores in calciumphosphate cement implants. J Mater Sci 2001;12:135–139.


160 LEUNG ET AL.

an in vitro optimized injectable calcium phosphate cement for augmenting screw fixation in...

Documents