INTRODUCTION

A variety of biological and synthetic materials areavailable for the surgical treatment of alveolar bonedefects. These are bone substitutes with variableabilities to induce bone regeneration and to sup-port proliferation of the bone-forming osteoblast-like cells. Bone substitute material is divided intotwo main families: bone derivative materials and in-organic calcium salts, subdivided into calciumphosphates and calcium carbonates.Autogenous bone grafts, which appear to be themost successful biological material, involve an addi-tional surgical procedure to harvest bone from anoral or extra-oral down site with associated risks andcomplications. Demineralized freeze-dried bone al-lografts (DFDBA) raise concerns about the consis-tency of results and the safety regarding infectious

disease transmission. An inorganic material derivedfrom mineralized bovine bone matrix is supposedto retain the physiological characteristics of bonemineral, but it raises some of the same concerns asDFDBA (1, 2).Calcium phosphate ceramics, mainly hydroxyap-atite (HA) and β-tricalcium phosphate (β-TCP), areused more than calcium carbonates in the peri-odontal field. Both HA, β-TCP and mixtures(biphasic calcium phosphates, BCP), have beenwidely used as calcified tissue substitutes in ortho-pedic treatments, alveolar ridge augmentation, pe-riodontal bone defects therapy, alveolar sockets andperi-implant fillings. Their chemical compositionsare close to the mineral fraction of bone, and theyhave a direct action on bone crystal formation (3,4). However, they have relatively little osteogenicproperties. Bioactive glass has been advocated for

Journal of Applied Biomaterials & Biomechanics 2003; 1: 186-193

Injectable calcium phosphate hydraulic cement(CPHC) used for periodontal tissue regeneration:A study of a dog model

F.J.G. CUISINIER1, A. WIEBER1, H. TENENBAUM1, P. VAN LANDUYT2, J. LEMAÎTRE2

1INSERM U595, Federation of Odontology Research, Department of Periodontology, Louis Pasteur University, 1Strasbourg - France2Laboratory of Poudres-EPFL Technology (LTP), Lausanne - Switzerland

ABSTRACT: Injectable calcium phosphate hydraulic cement (CPHC) is a new bone substitute family. This study aimed to eval-uate the use of CPHC in surgical periodontitis-simulating defects in a dog model. CPHC was obtained by adding powder mix-tures of different calcium phosphates with different solubility. Alveolar bone was removed by drilling over the mesial and distalroots of the 2nd mandibular premolar in six dogs. The defects were randomly selected, three were untreated and six treated. Thedefects had a depth of 6 mm and a width of 3 mm. The animals were sacrificed after 9 months and samples prepared, with nodecalcification, for histological evaluation.Seventy-nine percent of the root was covered by bone in the experimental defects, compared to 41% of the root for the control de-fects. Bone height was significantly higher for the experimental defects (4.9 ± 0.9 mm) than for the control defects (1.4 ± 0.5mm). After 9 months, 97 ± 6% of the CPHC was degradated and replaced by bone.This study proves the interest of this cement because of the particularly high level of periodontal bone regeneration. The abilityof the cement to be easily injected and shaped in bone defects and the immediate immobilization of the teeth after hardening isnotable. (Journal of Applied Biomaterials & Biomechanics 2003; 1: 186-93)

KEY WORDS: Injectable biomaterial, Calcium phosphate hydraulic cement, Animal study, Bone regeneration

Received 14/06/03; Revised 03/07/03; Accepted 15/09/03

1722-6899/186-08$15.00/0 © Società Italiana Biomateriali

187

Cuisinier et al

use in the same way as HA, β-TCP and BCP (5).None of these different synthetic materials reallysurpasses the others and their use is limited by theirlow initial mechanical stability mainly due to thepowder or particle structure.Injectable calcium phosphate hydraulic cement(CPHC) is a new bone substitute family. Biological(6) and biomechanical (7) properties of biodegrad-able CPHCs have been tested previously. The bio-logical responses to the CPHC were evaluated inboth close cylindrical defects in rabbit femurs (8, 9)and in large cylindrical defects in sheep (10). Thisstudy aimed to evaluate the use of such a CPHC inan experimental periodontal defect healing dogmodel. Therefore, the biological responses of thesurrounding tissues (cement, periodontal liga-ment, gingival tissues and alveolar bone), the re-sorption of the CPHC and the clinical relevance ofthe cement were investigated.

MATERIALS AND METHODS

CPHC

Individual CPHC doses of 1 ml were prepared bycombining powder mixtures of monocalcium phos-phate monohydrate (0.80 g, Ca(H2PO4)2.H2O,MCPM (Albright & Wilson)), β-tricalcium phos-phate (1.20 g, Ca3(PO4), β-TCP2, synthesized atLTP, batch LTP96-34) and dihydrogen sodium py-rophosphate (0.015 g, Na2H2P2O7 (Fluka)) with0.80 ml aliquots of 0.1 M H2SO4 aqueous solution(Merck p.a., diluted with ultra-pure water). Thepowders were sterilized by electron irradiation, andthe mixing liquid sterilized by autoclaving at 120 °Cfor 15 min.

Animals

Six adult female Labrador dogs (age: 9 ± 2 yrs) wereused in the study. Animal management, surgicalprocedures and routines were approved by the In-stitut National de la Santé et de la Recherche Med-ical (INSERM) (France), according to EuropeanCommunity guidelines for the care and use of lab-oratory animals (DE86/809/CEE).

Surgical procedures

Surgical procedures were performed under propo-fol general anesthesia (Diprivan®, Zeneca Pharma,France; 6 mg.kg–1.h–1). A broad-spectrum antibioticwas used for immediate post-surgical infection con-trol (Augmentin®, Smithkline Beecham, France; 2

g, intravenously). During the operating proce-dures, a ventilator with an endotracheal tube regu-lated the dogs’ breathing. Ketoprofene (Profenid®,Rhone Poulenc-Rorer, France; 100 mg, intra-venously) was used for immediate post-surgical paincontrol. Alveolar bone was removed by mechanicaldrilling on the mesial and distal roots of the 2ndleft mandibular premolar after buccal flap eleva-tion. The defects concerned only the buccal sur-faces of the tooth. A notch drilled in root dentinmarked the apical parts of the defects. The root sur-face was carefully planed and washed with distilledwater to eliminate bone debris. Buccal periodontaldehiscence treatments were randomly assigned asfollows: filled lesions (six), empty control lesions(three). The mean depth of experimental filled le-sions was 6.3 mm and their mean width was 3.3 mm.The mean depth and width of the control lesionswere 5.7 mm and 3.3 mm, respectively. These para-meters were measured with periodontal probes.The lesions were dried with a compress before fill-ing with CPHC for the experimental lesions or leftempty for the control lesions.The CPHC cement was applied with a syringe in thebone defect and pushed with a dry compress andshaped with a spatula before hardening. The le-sions were completely filled and the cement wasshaped to the previous alveolar bone crest. Thecoronal limits of cements corresponded to the alve-olar crest. The hardening time (10 min) was repro-ducible and allowed good management of theCPHC.After shaping and cement hardening, the flaps,which were immobilized by non-resorbable sutures,covered the lesions.

Samples harvesting

The dogs were sacrificed after 9 months using asodium pentobarbital overdose (Dolethal®, Ve-topharma, France) during a general anesthesia in-duced by an intramuscular injection of 200 mg ofketamine chlorhydrates (Ketalac®, Parke-Davies,France). The premolars and the surrounding tis-sues were collected by section of the mandibularbone.

Specimen preparation

After fixing for 7 days in a 2% glutaraldehyde and2% paraformaldehyde solution in 0.1 M sodium ca-codylate buffer, pH 7.4, the specimens were em-bedded in Epon. After hardening, blocks were cutwith a diamond saw (Escil, France) at a thickness of0.1-0.3 mm. After grinding and polishing, the sec-

188

Injectable CPHC used as periodontal tissue regeneration

tions were glued to a plexiglass plate with cyanocry-late glue (Superbond, DuPont, USA) and then col-ored by Goldner-Masson trichrome.

Observations and measurements

Optical microscopic images (Zeiss, Jena, Germany)were recorded either on photographic film or di-rectly with a color CCD camera (Sony, New Jersey,USA) connected via a frame grabber board (MiroDC30, Pinnacle System, USA) to a pentium basedPC (ABC, Strasbourg, France). Measurements were collected on numerical imageswith the NIH image software (Scion Corporation,USA).

Quantitative evaluation

The following parameters, defined according to the AS-BMR histomorphometry nomenclature, were used:– height of the new bone measured from the bot-

tom of the notch to the alveolar bone crest (BH);– distance between the alveolar bone crest to the

cemento-enamel junction (bone-CEJ); – ratio between the BH and the sum of BH and

bone-CEJ (BH/root); – gingival height measured from the alveolar bone

crest to the gingival crest (gingival height);– ratio of CPHC surface to bone surface (CPHC/

bone).The BH/root ratio gave a good indication of theavailable fraction of root covered by the newlyformed bone. The CPHC/bone ratio was an indi-cator of cement degradation.

Statistics

The significance of the differences between the av-erage responses of the experimental and the con-trol groups were assessed using the Student’s pairedt-test, and expressed in terms of Type-I error risks(p, probability of the observed difference to occurby pure chance). For statistical normalization pur-poses, the responses analysed were: log(BH),log(BS), log(bone-CEJ) and log[1/(root/BH-1)].Outliers were discarded using the Student’s t-teston the transformed responses (p < 0.10).

RESULTS

Clinical observations

The clinical aspect of the gingiva at the lesion sitewas carefully noted before sample harvesting. The

gingival height was similar for experimental andtest lesions (Fig. 1) and there were no clinical signsof gingival inflammation.

Histological observations

In three experimental lesions, small CPHC parti-cles were observed (Fig. 2). In all the defects, thenew alveolar bone showed normal histological or-ganization. The root dentin, which was carefullyplaned during surgery, was partially covered withthe new cementum when cementogenesis occurred(Fig. 3). The extent of the root covered by the newcementum was limited. In the coronal part of thedefect, root dentin was lined directly by the newalveolar ligament (Fig. 4). In Figures 3 and 4, thenew periodontal ligament is clearly visible and wenever observed ankylosis. The ligament width andstructure were similar in the upper or lower portion(apical to the notch) of the root and the collagenfibers had a normal oblique orientation with denti-nal insertion apical to bone insertion (Fig. 4). The resorption mechanism of the CPHC was an ac-tive cellular process, as shown in Figure 5. The os-teon was created by a first osteoclastic resorption.During this stage, CPHC resorption was performedin same time than bone as indicated by the curvedaspect of the resorpting front in the CPHC. Os-teoblast cells then deposited circular layers of newbone. These layers had a deeper red color than theoriginal bone. In the root portion, not covered by ligamentand bone, dentin was directly in contact with theconjunctive gingival tissue. In the case of the control defects, with noCPHC, root dentin resorption was observed(Fig. 6) and was indicated by the presence of re-sorption lacunae located on the dentin surface.These resorption lacunae were not seen in thecase of the experimental defects filled withCPHC.

Quantitative evaluation

The results of the quantitative comparison be-tween CPHC-treated and control defects arepresented in Figure 7. Bone height (Fig. 7a) andbone surface (Fig. 7b) are significantly higherfor the CPHC-treated defects (p < 0.05 and0.0005, respectively). The distance between thealveolar bone crest to the bone-CEJ and the ra-tio between bone height and available root(BH/root) were determined as being signifi-cantly different between the two groups. Theformer ratio can be seen as the percentage of

189

Cuisinier et al

Fig. 1 - View of a premolar immediately after dog sacrifice. Theabsence of gingival inflammation and the regular gingival mar-gin are noted.

Fig. 2 - Histological section in the apical part of an experimentaldefect after 9 months of healing. CPHC: non degraded CPHC; N:notch indicated apical part of experimental lesion. Note the newbone formation coronal to the experimental notch.

Fig. 3 - Higher magnification view of the notch presented in Fig-ure 2. The new cementum is clearly visible in contact with the rootdentin. PL: periodontal ligament.

Fig. 4 - Coronal portion of an experimental defect without ce-mentogenesis; collagen fibers (CF) of the periodontal ligament(PL) are directly inserted in dentin.

190

Injectable CPHC used as periodontal tissue regeneration

available root covered by new bone. For experi-mental defects filled with CPHC, 79% of avail-able root was covered by bone, compared to41% for control defects (Fig. 7c, p < 0.02) Thebone-CEJ distance was also significantly smallerfor CPHC-treated defects compared to controls(Fig. 7d, p < 0.02). The cement height for experimental defects wasrather high with a significant standard devia-tion. The difference between the two groups wasnot significant. However, it appears that in anunpredictable number of CPHC cases the ce-ment regeneration was very high. The new bone surface was 4.3 ± 0.8 mm2 for de-fects filled with CPHC and 0.5 ± 0.1 mm2 forcontrol defects. In the CPHC filled defects, thenon-degraded cement represented a surfacearea of 0.15 ± 0.2 mm2. The ratio between theCPHC surface and the bone surface was restrict-ed to the experimental group. This ratio indi-cated that 97 ± 6% of the CPHC was degradedand replaced by bone, assuming that initialCPHC volume was equal to the newly formedbone volume. This assumption is because duringsurgery CPHC was easily shaped to the alveolarbone morphology. In four cases, the CPHC wascompletely resorbed.

DISCUSSION

The dog model used in this study was based on sur-gical periodontitis-simulating defects. The meansize of the defects, measured with periodontalprobes was close to the defect height measuredfrom biopsies on histological preparation. The

mean defect height was homogeneous, as request-ed, to allow a reliable histometrical analysis (11).The experimental surgically created defects wererandomly assigned to be filled or not with CPHC,but the number of control defects was limited tothree. If we compare our experimental set up withpreviously published studies of regenerative andbone grafting periodontal surgery in dog models,the main difference was in the healing time. Proto-cols involved dog sacrifices at 4 or 8 weeks (11, 12).Such a relatively short healing time allowed clearproof of biological cytokine activities, but a 4-weekhealing interval appears insufficient for assessingDFDBA integration (12). When periodontal regen-eration depends on the degradation speed of a ce-ramic (i.e. HA on coralline implant) the biopsiesare mainly collected at 4, 6 or 8 months and even

Fig. 5 - Bone-CPHC interface. RF: resorption front; HC: vascularcanal of osteon.

Fig. 6 - Coronal part of a control defect. RL: dentin resorption la-cunae.

191

Cuisinier et al

later (13-15). This literature survey shows that ourstudy is usefully comparable to previous dog stud-ies. The good biocompatibility of this CPHC cement wasproven in different animal studies (5-7) and is con-firmed by our study. The cement degradation was al-most complete after 9 months. This is partially dueto the anatomy of the periodontal defect. The average bone regeneration height (Fig. 7a) was4.9 ± 0.9 mm for the CPHC group and 1.4 ± 0.01mm for the control defects (95% confidence inter-

vals). This value could be compared to bone regen-eration obtained in similar animal studies. For tran-forming growth factor-beta 1 (TGF β1) and recom-binant human bone morphogenetic protein-2(BMP-2), the bone height of the experimental de-fects vs the control defects were 1.8 vs 1.3 mm and3.5 vs 0.8 mm, respectively, after 8 weeks of healing(16, 17). The relatively short regeneration time ex-plains the low bone growth for the control defectscompared to the 1.4 mm we obtained. Bone heightsfor the experimental defects were smaller in the

Fig 7 - Quantitative comparison between CPHC-treated and control defects: a) Bone height (BH, mm); b) Bone surface (BS, mm2); c) Bone crest height/total root height ratio (BH/root, %); d) Distance from bone to cement-enamel junction (bone-CEJ, mm).

a b

dc

Adj. Values

Exp. Values

p < 0.005

p < 0.02 p < 0.02

p < 0.0005+3 S.E.

-3 S.E.

192

Injectable CPHC used as periodontal tissue regeneration

previous studies than the 4.9 mm we obtained. Af-ter 8 weeks of healing, the bone height regeneratedwas 4.0 mm for DFDBA, but large areas of the graftwere not resorbed (18). Comparatively, after 9months a quasi-total CPHC biodegradation and astatistical difference for bone height between testand control defects were achieved.Our work is also similar to a study in which bioac-tive glass and DFDBA were compared in the treat-ment of defects around titanium implants (14). ForDFDBA 65.7% of the bone defect was filled and forthe control defects 48% of bone defect was regen-erated. For bioactive glass filled defects and controldefects no significant difference was observed after3 months. For CPHC filled defects 79.2 ± 11.4% ofthe bone height regenerated and for control site re-generation was averaged at 41 ± 13.5%, which is sur-prisingly lower than for the defects realized in con-tact with titanium implants. The cementogenesis, determined by measuring theheight of the cement growth from the apical part ofthe notch, was not statistically different between thetwo groups. However, we made two observations.The non-statistical difference was related to thehigh standard deviation of cement height in the ex-perimental defects. 50% of the defects showed nocementogenesis, but in 50% a very high growth wasnoted. Secondly, the cement growth for the controldefects was lower, as expected from previous guid-ed tissue regeneration (GTR) studies (19, 20). Thisindicates that in our experiment the cementgrowth was unpredictable, but also that the CPHCgraft could protect precursor cementoblast cellsand could not suppress or perturb cellular migra-tion towards the dentin surface. Osteogenesis, assessed by different parameters(bone surface, bone-CEJ distance, BH:bone ratio),was significantly higher in the test specimens than incontrols. Periodontal ligament regeneration wascompleted with collagen fibers running from thealveolar bone surface to the root surface.The two cellular populations, which are also con-cerned by the CPHC graft are gingival fibroblastsand junctional epithelium cells. The careful exami-nation of the root surface in the portion betweenthe new bone crest and junctional epithelium showfor control defects resorption lacunae. Such re-sorption lacunae are, according to early GTR stud-ies, created by gingival fibroblasts (19). The ab-sence of resorption lacunae in grafting defects in-dicates that CPHC prevents the root contaminationby fibroblast cells. This indicates that CPHC has aprotective effect related to membranes in the GTRtechnique. This was related to the bonding betweenCPHC and root dentin, during cement hardening.

As noted, CPHC degradation is due to active cellu-lar events similar to the ones occurring for boneturnover. The mononuclear precursors of osteo-clast-like cells reach the resorption area via capil-lary action. In our experiment, they came mainlythrough the alveolar bone capillaries. Therefore,the dentin surface is the last surface liberated fromCPHC. Kinetical studies are needed to understandfully the CPHC degradation process when used as aperiodontal graft.This study proves the interest of this cement for thetreatment of periodontal defects because of theparticularly high level of periodontal tissue regen-eration and of a possible GTR-like effect. The abili-ty of the cement to be easily injected and shaped inbone defects and the immediate immobilization af-ter cement hardening also favor CPHC use in peri-odontal treatments.

ACKNOWLEDGEMENTS

PVL and JL wish to acknowledge the support of the Board of theSwiss Federal Institutes of Technology for supporting this work(PPM project n° 4.2D), and also the support of the RobertMathys Foundation and of SRATEC Medical.HT and FJGC acknowledge the assistance of Dr. Jean-Francois Schaff, Dr. Hubert Gros and Dr. DominiqueRoth during surgery and the technical help of Dr. AlinaGamelescu and Dr. Stéphane Duisit during specimenpreparation.

Address for correspondence: Dr. Frédéric Cuisinier INSERM U59511 rue Humann 67085 Strasbourg Cedex France fred.cuisinier@odonto-ulp.u-strasbg.fr

193

Cuisinier et al

REFERENCES

1. Camelo M, Nevius M, Schenk R. Clinical radi-ographic and histologic evaluation of human peri-odontal defects treated with Bio-oss and Bio-glide.Int J Periodontics Restorative Dent 1998; 18: 321-31.

2. Bowen JA, Mellonig JT, Gray JL, Towle HT. Comparisonof decalcified freeze-dried bone allograft and porousparticulate hydroxyapatite in human periodontal os-seous defects. J Periodontol 1989; 60: 647-54.

3. Rey C. Calcium phosphate biomaterials and bonemineral. Differences in composition, structures andproperties. Biomaterials 1990; 11: 13-5.

4. Hemmerle J, Cuisinier FJG, Schultz P, Voegel JC.HRTEM study of biological crystal growth mecha-nisms in the vicinity of implanted synthetic hydrox-yapatite crystals. J Dent Res 1997; 76: 682-7.

5. Neo M, Kotani S, Nakamura T, et al. A comparativestudy of ultrastructure of the interfaces between fourkinds of surface-active ceramic and bone. J BiomedMater Res 1992; 26: 1419-32.

6. Ohura K, Bohner M, Hardouin P, Lemaitre J,Pasquier G. Resorption of, and bone formationfrom, new b-tricalcium phosphate: monocalciumphosphate cements: An in vivo study. J Biomed MaterRes 1995; 30: 193-200.

7. Ikenaga M, Hardouin P, Lemaitre J, Andrianjatovo H,Flautre B. Biomechanical characterization of abiodegradable calcium phosphate hydraulic cement: acomparison with porous biphasic calcium phosphatecement. J Biomed Mater Res 1998; 40: 1393-400.

8. Pasquier G, Flautre B, Blary MC, Anselme K,Hardouin P. Injectable percutaneous bone biomate-rials. An experimental study in a rabbit model. JMater Sci Mater Med 1996; 7: 683-90.

9. Pasquier G, Flautre B, Leclet H, Hardouin P. Experi-mental evaluation of a percutaneous injectable bio-material used in radio-interventional bone fillingprocedures. J Mater Sci Mater Med 1998; 9: 333-6.

10. Flautre B, Delecourt C, Blary MC, Van Landuyt P,Lemaitre J, Hardouin P. Volume effect on biologicalproperties of a calcium phosphate hydraulic cement:Experiment study in sheep. Bone 1999; 25 (suppl 2):S 35-9.

11. Haney JM, Zimmerman GJ, Wikesjo UM. Periodon-tal repair in dogs: evaluation of the natural diseasemodel. J Clin Periodontol 1995; 22: 208-13.

12. Caplanis N, Lee MB, Zimmerman GJ, Selvig KA,Wikesjo UM. Effect of allogenic freeze-dried dem-ineralized bone matrix on regeneration of alveolarbone and periodontal attachment in dogs. J Clin Pe-riodontol 1998; 25: 801-6.

13. Moon IS, Chai JK, Wikesjo UM, Kim CK. Effects ofpolyglactin mesh combined with resorbable calciumcarbonate or replaminform hydroxyapatite on peri-odontal repair in dogs. J Clin Periodontol 1996; 23:945-51.

14. Hall EE, Meffert RM, Hermann JS, Mellonig JT,Cochran DL. Comparison of bioactive glass to dem-ineralized freeze-dried bone allograft in the treat-ment of intrabony defects around implants in the ca-nine mandibule. J Periodontol 1999; 70: 526-35.

15. West TL, Brustein DD. Freeze-dried bone andcoralline implants compared in the dog. J Periodon-tol 1955; 56: 348-51.

16. Wikejso UM, Razi SS, Sigurdsson TJ, et al. Periodon-tal repair in dogs: effect of recombinant humantransforming growth factor-beta 1 on guided tissueregeneration. J Clin Periodontol 1998; 25: 475-81.

17. Sigurdsson TJ, Lee MB, Kubota K, Turek TJ, WozneyJM, Wikejso UM. Periodontal repair in dogs: recom-binant human bone morphogenetic protein-2 signif-icantly enhances periodontal regeneration. J Peri-odontol 1995; 66: 131-8.

18. Kim CK, Cho KS, Choi SH, Prewett A, Wikesjo UM.Periodontal repair in dogs: effects of allogenicfreeze-dried demineralized bone matrix implants onalveolar bone and cementum regeneration. J Peri-odontol 1998; 69: 26-33.

19. Warrer K, Karring T. Guided tissue regenerationcombined with osseous grafting in suprabony peri-odontal lesions. An experimental study in dog. J ClinPeriodontol 1992; 19: 373-80.

20. Sigurdsson TJ, Harwick R, Bogle GC, Wikejso UM.Periodontal repair in dogs: space provision by rein-forced ePTFE membranes enhances bone and ce-mentum regeneration in large supraalveolar defects.J Clin Periodontol 1998; 25: 475-81.