IOP PUBLISHING BIOMEDICAL MATERIALS

Biomed. Mater. 4 (2009) 015012 (6pp) doi:10.1088/1748-6041/4/1/015012

Viability of osteocytes in bone autograftsharvested for dental implantology

Bernard Guillaume1, Christine Gaudin2, Sonia Georgeault2,Romain Mallet2, Michel F Basle2 and Daniel Chappard2

1CFI—College Francais d’Implantologie, 6 rue de Rome, 75008 Paris, France2 INSERM, U 922—LHEA, Faculte de Medecine, 49045 Angers Cedex, France

E-mail: daniel.chappard@univ-angers.fr

Received 2 September 2008Accepted for publication 17 November 2008Published 12 December 2008Online at stacks.iop.org/BMM/4/015012

Abstract

Bone autograft remains a very useful and popular way for filling bone defects. In maxillofacialsurgery or implantology, it is used to increase the volume of the maxilla or mandible beforeplacing dental implants. Because there is a noticeable delay between harvesting the graft andits insertion in the receiver site, we evaluated the morphologic changes at the light andtransmission electron microscopy levels. Five patients having an autograft (bone harvestedfrom the chin) were enrolled in the study. A small fragment of the graft was immediately fixedafter harvesting and a second one was similarly processed at the end of the grafting periodwhen bone has been stored at room temperature for a 20 min ± 33 s period in saline. A netincrease in the number of osteocyte lacunae filled with cellular debris was observed (+41.5%).However no cytologic alteration could be observed in the remaining osteocytes. The viabilityof these cells is known to contribute to the success of autograft in association with other lesswell-identified factors.

(Some figures in this article are in colour only in the electronic version)

Introduction

Materials for replacing bone are necessary in a number ofreconstruction surgeries of the skeleton caused by traumatic,tumor-resection or congenital defects. The highest demandconcerns orthopedic surgery for the treatment of femoralprosthesis loosening where wear debris induces a markedosteolysis. It has been estimated that more than 500 000bone grafts are done each year in the USA [1]. However,there is also a growing demand in maxillofacial surgeryand in implantology to restore a sufficient volume of bonebefore placing implants in the mandible or the maxillar[2]. Allogenic bone grafts are common in orthopedics butrequire access to a bone bank in some countries [3]. Thehigh cost of a bone bank, the difficulties of conditioningbone chips of fragments specially adapted to maxillofacialsurgery and the potential risk of transmission of viral or priorinfections have made allografts seldom used in implantology.Synthetic phosphocalcic materials (β-tricalcium phosphate,hydroxyapatite) are numerous in the market but are associated

with various biomechanical and osteoconductive properties.None of them have been found sufficient to restore large bonedefects. Injectable calcium phosphate materials have beenproposed but their long-term effects could not be detectablein a multicentric orthopedic study [4]. In a recent animalstudy conducted in this laboratory, an injectable bone pastebased on βTCP-hydroxyapatite induced osteoconduction withwoven bone apposition but the remodeling process removedit in a second time, making the rationale of such substitutesquestionable [5]. In addition, synthetic ceramics are often toobrittle to be used in weight-bearing bones or when compressivestresses are too high.

In orthopedics, bone is commonly harvested from theiliac crest, as it provides easy access to good quality andquantity cancellous autograft. Other sources are Gerdy’stubercle of the tibia and the distal parts of the radius or tibiafor cancellous bone; autologous cortical bone can be obtainedfrom the fibula [6], ribs and iliac crest. These sites werealso widely used in maxillofacial surgery from as early asWorld War II [7]. However other sites have been favored

1748-6041/09/015012+06$30.00 1 © 2009 IOP Publishing Ltd Printed in the UK

Biomed. Mater. 4 (2009) 015012 B Guillaume et al

for the reconstruction of maxillofacial bone defects such asparietal [8] or mandibular bone [9]. These observationswere based on experimental works in animals showing thatmembranous bone grafts survive better than endochondralbone after grafting [10, 11]. Parietal or chin bone grafting hasbeen proposed in implantology because access is simpler in thesurgical theatre than for other sites. Autologous bone providessuperior results to other methods since it is revascularizedeasily and rapidly incorporated into the recipient site. It isalso commonly thought that osteoblasts, osteocytes (OCs) andlining cells of the graft can survive the transplant and favorthe osseointegration of the graft [12]. However, this hasbeen seldom documented in the literature. This study wasundertaken to survey the ability of bone cells to survive duringthe transplant process starting from harvesting to implantation.

Material and methods

Patients and surgical protocol

Human mandibular bone samples were obtained from fivepatients under general anesthesia during the time course ofgrafting for pre-implantation. Each patient has given his/herinformed consent to participate in the present study. Thesurgical protocol aims at increasing the bone wall thicknessat the maxilla or mandible in deficient recipient sites priorto the placement of dental implant(s) of standard diameter.A broader bone volume ensures an easier position for theimplant’s axis. The grafting zone was operated on first toappreciate and determine the shape, volume and position ofthe graft.

For each patient, the graft was harvested at the chin;a preliminary radiographic assessment has been done tocheck the axis and the thickness of the donor site. One ortwo rectangular cortico-cancellous bone samples and cortico-cancellous cylinders were removed. The intra sulcular incisionwas made at the level of the dental collars with two sideincisions at the mesial first premolars. This full thicknessreflection flap is preferred to an incision at the gingival marginbecause it allows a better coverage flap of the grafting zonewithout any tension and exposure of the graft (figure 1). Thelimits of the harvesting graft areas were done with a thin burfor bone to outline the defect margins on the buccal bone plateto a 6 mm depth. The harvested graft was separated with achisel by progressive cleavage. Chips of cortico-cancellousbone (mostly cortical) were collected. The bone sampleswere immediately placed in sterile saline at room temperatureuntil used for grafting. One small sample (1 mm × 1 mmfrom the chin cortex and containing the endosteal surface) wasimmediately fixed in a glutaraldehyde-based fluid. The marginflaps were closed with interrupted sutures with 5/0 vycril. Foreach patient, the amount of the graft corresponded roughly to1.8 cm3 per cortico-cancellous block graft (usual dimensions:10 mm × 6 mm × 3 mm). The flow chart of the study appearsin figure 2.

When harvesting bone was achieved, the surgeon began torelease the muccoperiosteal flap; the margin recipient bone sitewas revived with a round bur to obtain a slightly hemorrhagiczone. The site was carved to create a mortise favorable so

(A)

(B)

Figure 1. Harvesting the bone graft at the chin. (A) Section of bonewith the thin bur. (B) Mobilization of the cortico-cancellous graftswith a chisel.

Figure 2. Flow chart of the study from harvesting to grafting. Thespecimens collected for histological analysis were takenimmediately and at the end of the grafting period.

to create a primary closure and the anchorage of the graftwas then stabilized with stainless steel screws (the cancellousbone of the cortico-cancellous harvested graft was placed indirect contact with the patient’s recipient site) (figure 3(A)).The cancellous bone chips were placed on all sides around

2

Biomed. Mater. 4 (2009) 015012 B Guillaume et al

(A)

(B)

Figure 3. (A) The cortico-cancellous grafts are immobilized in therecipient site by screwing. (B) Clinical aspect of a well-integratedallograft at 6 months.

and on top of the monocortical block graft. The full thicknessflap was then closed to the primary incisions and sutured with5/0 vicryl. When the last chip was positioned, the surgeonplaced a remaining bone sample into a new vial containingthe glutaraldehyde fixative. In this way, the effects of storingin physiological saline at room temperature during a mean of20 min ± 33 s could be explored on paired specimens. Sutureswere removed two weeks post-operatively. A CT-scan wasperformed at 3 and 6 months post-surgery to ensure the bonegraft healing. The graft was deemed successful when a densebony reaction could be seen at the grafted site.

Bone microscopy

The fixative was prepared in ready-to-use plastic vials thatwere stored in the refrigerator until use. The fixative wascomposed of glutaraldehyde 4% in cacodylate buffer, pH 7.4.Samples were fixed for 1 h at 4 ◦C in the surgical unit thenrinsed, stored in cacodylate buffer and sent to the laboratory.The samples were then cut with a clean blade into smallpieces suitable for electron microscopy and post-fixed in 1%osmium tetroxide. They were dehydrated through a gradedethanol series and finally embedded in Epon 812 according tostandard methods used for TEM. The blocks were stored atroom temperature until ready to use.

For each block, five sections (1 μm in thickness)were obtained semi-serially every 20 μm (this separationensures that no osteocyte lacuna profile could appear ontwo consecutive sections). A Leica Ultracut S (Leica–Rueil Malmaison, France) was used with a glass knife andsections were immediately deposited on a glass slide with adrop of distilled water. The sections were dried at 60 ◦C;then they were incubated for 15 min in periodic acid44 mM and stained with an extemporaneously preparedsolution combining aqueous methylene blue 1% (1 volume)and azure II 1% (1 volume) for 20 min at room temperature.The sections were then thoroughly washed in distilled waterand mounted in NeoEntellanTM (Merck) after air drying.

On each set of sections prepared per block, the totalnumber of osteocyte lacunae was counted by light microscopyon the five sections. Lacunae were separately classified ascontaining a normal OC with nucleus and cytoplasm withprolongations, empty lacunae and lacunae containing cellulardebris. The results were expressed as a percentage of the totallacunae.

Transmission electron microscopy

Ultrathin sections (80 nm in thickness) were obtained with adiamond knife and transferred onto nickel grids (mesh 100)coated with a film of collodion. Grids were then air-dried,contrasted with sodium metaperiodate, uranyl acetate and leadcitrate. Observations were done with a transmission electronmicroscope JEOL 2011 (JEOL-France) at 120 kV.

Statistical analysis

Statistical study was done using Systat 11 (Systat softwareInc.). Data were expressed as mean ± standard error of themean (SEM). Significant differences between samples wereassessed with nonparametric Mann and Whitney’s U test.Differences were considered as significant when P < 0.05.

Results

The patients were re-examined in the days following the graft,even in the absence of any complication. The absence ofprolonged pain, edema, purulent discharge, elimination ofsequestrated and devitalized bone fragments in the followingmonths are in favor of a progressive success of graft fixation.The palpation of the grafted site showed an increased, stableand painless relief (figure 3(B)). Radiographs (especially theCT-scans) confirmed adherence of the graft at the receiver siteby the absence of radiolucent space between the two zones.The gain in thickness was 5 mm on average and 4 mm implantscould be placed with a satisfactory axis.

The methylene-blue-azure staining allowed a clear-cutidentification of the various types of lacunae. OCs were clearlyevidenced by a blue cytoplasm with processes extending intothe matrix lacunae. Empty lacunae were devoid of any cellularmaterial while cellular debris could be identified inside or onthe margins of the lacunae and corresponded to remnants ofnecrotic cells. The matrix itself was unstained, or appearedlight bluish (figure 4).

3

Biomed. Mater. 4 (2009) 015012 B Guillaume et al

(A)

(B)

(C)

Figure 4. The different types of osteocyte lacunae in lightmicroscopy. (A) Lacuna filled with a typical osteocyte with awell-defined nucleus. (B) Lacuna containing cellular debris.(C) Empty lacuna. Original magnification ×1000.

OC counts appear in figure 5; counts are presented on thesamples studied at the beginning and the end of the graftingsurgery. The number of empty lacunae remained unchangedafter the period spent in ex vivo conditions. On the other hand,a significant increase in the number of lacunae containingdebris was observed (+41.5%) and conversely, the number oflacunae containing an intact osteocyte was reduced.

On TEM sections, the OC lacunae were observed and theanalysis of the fine cytological details of these cells did notreveal any change (figure 6). Occasionally some osteoblastswere encountered and no gross abnormalities (e.g., recyclingmembrane) were evidenced.

Figure 5. OC count on semi-thin sections. (A) The number ofosteocyte lacunae filled with cell debris increased in all patients.(B) Cumulative histogram of the three different types of lacunae atthe beginning and the end of the grafting surgery.

Discussion

Cell necrosis can occur rapidly when cells are deprivedof oxygen and essential metabolites such as glucose [13].The process differs from apoptosis, which is a particularmechanism induced by various factors (excess or deprivationof hormones, growth factors or cytokines, p53 . . . ). Necrosis,unlike apoptosis, can induce an inflammatory reaction.However, osteoinductive factors released from the graft duringthe resorptive process, as well as cytokines released locally bythe inflammatory phase, are thought to contribute to healingof the graft. It is likely that other factors may play akey role in the excellent success rate of autografts becausedevitalized allografts can release similar molecules. Thepresence of osteogenic cells is frequently advocated as thebasis requirement for the success of autografts [12]. Corticalbone has a higher fractional bone volume (i.e., percentageof the tissue volume occupied by bone) than cancellousbone and has a lower porosity. Haversian canals are themain reservoir of osteogenic cells together with the endostealsurfaces but the number of osteoblast precursors is reduced

4

Biomed. Mater. 4 (2009) 015012 B Guillaume et al

(B)

(A)

(C)

Figure 6. Transmission electron microscopy of living osteocytes ina graft sample at the beginning (A) and the end (B) of the surgicalprocedure. A group of osteoblasts is also shown at the end (C).Scale bars: 2 μm.

when compared to cancellous bone [14]. Also, the absenceof hematopoietic tissue is recognized to provide fewer stemcells in cortical bone autografts. In orthopedic surgery,cortical bone grafts are favored in loaded areas and they havebeen found to be more resistant to vascular ingrowth andremodeling [15]. Vascularized cortical grafts (i.e., parts ofharvested bone with their vascular pedicle, reanastomosedat the recipient site) have been proposed to provide bonewith viable cells and restore a local vascularization inside thegraft [16]. Although the method has shown clearly superiorresults with large bone defects (e.g. >12 cm), it is associatedwith increased morbidity. The continued vascular supplyis thought to allow for faster bone incorporation but clearhistopathologic analyses are lacking. To our knowledge,

no data have been presented on the potential changes inviability of bone cells induced by the harvesting and storageprocedures used before grafting. Another way to evaluate theviability of the osteocytes in the bone chips would have beento perform a LDH (lactate dehydrogenase) histoenzymaticanalysis. However in the present study, the grafts wereharvested far from the laboratory and we chose to evaluate bonecell preservation by using morphological EM criteria (absenceof mitochondrial or membrane changes). Furthermore,the method necessitates cryosectioning, which is notcompatible with EM techniques [17]. In the present study,the number of osteoblasts that could be seen on the corticalmorcellized chips was very low, and only OCs were evidencedin all subjects. OCs have a reduced number of organiteswhen compared to osteoblasts. Because OCs reside distantfrom the blood supply, their metabolic needs are satisfiedby a combination of passive diffusion of fluids through thematrix and canaliculi and enhanced diffusion arising when theskeleton is loaded during functional activity [18–20]. Theirmetabolic demand is probably lower than other high energyspending cells and this can explain why a large fraction of OCssurvived the surgical conditions (room temperature, saline,absence of oxygen and metabolites). In normal humans, anoticeable fraction of lacunae are found empty, the fractionis increased in osteoporotic patients [21]. OCs are highlydifferentiated cells which differ from osteoblasts [22, 23]. Invitro studies on the MLO-Y4 osteocytic cell line have foundthe possibility of these cells to dedifferentiate into osteoblasts[22]. Furthermore, the importance of OCs in the successof bone autografting was stressed in a recent animal study:bone-grafted particles healed when living OCs were presentand failed when OCs have undergone necrosis [24]. Anotherimportant factor that could explain the superiority of autograftversus allograft is the total absence of protein denaturation inthe former. The 20 min incubation in saline used in this study isprobably harmless to the patient’s proteins. On the other hand,protein denaturation is known to occur after deep freezing [25]or autoclaving [26]. Treatment of bone allografts with variouschemical processes used to clean bone was recently found toalter some important matrix proteins [27].

This study is the first to report the cellular effects ofthe limited storage conditions used surgically to preservemorcellized bone chips of the chin. A noticeable proportionof OCs remains healthy without modifications observed at theelectron microscopy level; other OCs died by necrosis. Thesuccess of autografts could result in the preservation of livingOCs together with additional factors such as better proteinpreservation.

Acknowledgment

The authors thank Mrs Laurence Lechat for secretarialassistance. This work was supported by funds from ‘Paysde la Loire–Bioregos’ and INSERM.

References

[1] Bucholz R W 2002 Nonallograft osteoconductive bone graftsubstitutes Clin. Orthop. Relat. Res. 395 44–52

5

Biomed. Mater. 4 (2009) 015012 B Guillaume et al

[2] Giannoudis P V, Dinopoulos H and Tsiridis E 2005 Bonesubstitutes: an update Injury 36 (Suppl 3) S20–7

[3] Tomford W W and Mankin H J 1999 Bone banking. Update onmethods and materials Orthop. Clin. North Am. 30 565–70

[4] Finkemeier C G 2002 Bone-grafting and bone-graft substitutesJ. Bone Joint Surg. Am. A 84 454–64

[5] Blouin S, Moreau M F, Weiss P, Daculsi G, Basle M F andChappard D 2006 Evaluation of an injectable bonesubstitute (betaTCP/hydroxyapatite/hydroxy-propyl-methyl-cellulose) in severely osteopenic and aged ratsJ. Biomed. Mater. Res. A 78 570–80

[6] Askew C A, Fleming P S and Downie I P 2005 Adjunctive useof dental implants in mandibular reconstruction with a freefibular graft Br. J. Oral Maxillofac. Surg. 43 341–2

[7] Abubaker A O, Sotereanos G C and Patterson G T 1992Cranial bone harvesting Modern Practice in Orthognathicand Reconstructive Surgery ed W Bell (Philadelphia, PA:Saunders) pp 975–84

[8] Jung Y S, Kim H J, Choi S W, Kang J W and Cha I H 2003Regional thickness of parietal bone in Korean adults Int. J.Oral Maxillofac. Surg. 32 638–41

[9] Widmark G, Andersson B and Ivanoff C J 1997 Mandibularbone graft in the anterior maxilla for single-tooth implantspresentation of surgical method Int. J. Oral Maxillofac.Surg. 26 106–9

[10] Smith J D and Abramson M 1974 Membranous vsendochondrial bone autografts Arch. Otolaryngol 99 203–5

[11] Zins J E and Whitaker L A 1983 Membranous versusendochondral bone: implications for craniofacialreconstruction Plast. Reconstr. Surg 72 778–85

[12] Burchardt H 1983 The biology of bone graft repair Clin.Orthop. Relat. Res. 174 28–42

[13] Hlatky L, Sachs R K and Alpen E L 1988 Jointoxygen-glucose deprivation as the cause of necrosis in atumor analog J. Cell. Physiol. 134 167–78

[14] Manolagas S C and Jilka R L 1995 Bone marrow, cytokines,and bone remodeling emerging insights into thepathophysiology of osteoporosis N. Engl. J. Med.332 305–11

[15] Boden S D and Schimandle J H 1995 Biologic enhancement ofspinal fusion Spine 20 (S24) 113–23

[16] Enneking W F, Eady J L and Burchardt H 1980 Autogenouscortical bone grafts in the reconstruction of segmentalskeletal defects J. Bone Joint Surg. Am. 62 1039–58

[17] Wong S Y, Dunstan C R, Evans R A and Hills E 1982 Thedetermination of bone viability: a histochemical method foridentification of lactate dehydrogenase activity inosteocytes in fresh calcified and decalcified sections ofhuman bone Pathology 14 439–42

[18] Jensen S S, Broggini N, Hjorting-Hansen E, Schenk R andBuser D 2006 Bone healing and graft resorption ofautograft, anorganic bovine bone and beta-tricalciumphosphate A histologic and histomorphometric studyin the mandibles of minipigs Clin. Oral Implants Res.17 237–43

[19] Bonewald L F 2002 Osteocytes: a proposed multifunctionalbone cell J. Musculoskelet. Neuronal Interact. 2 239–41

[20] Takai E, Mauck R L, Hung C T and Guo X E 2004 Osteocyteviability and regulation of osteoblast function in a 3Dtrabecular bone explant under dynamic hydrostatic pressureJ. Bone Miner. Res. 19 1403–10

[21] Qiu S, Rao D S, Palnitkar S and Parfitt A M 2003 Reduced iliaccancellous osteocyte density in patients with osteoporoticvertebral fracture J. Bone Miner. Res. 18 1657–63

[22] Kato Y, Windle J J, Koop B A, Mundy G R and Bonewald L F1997 Establishment of an osteocyte-like cell line, MLO-Y4J. Bone Miner. Res. 12 2014–23

[23] Barragan-Adjemian C, Nicolella D, Dusevich V, Dallas M R,Eick J D and Bonewald L F 2006 Mechanism by whichMLO-A5 late osteoblasts/early osteocytes mineralize inculture: similarities with mineralization of lamellar boneCalcif. Tissue Int. 79 340–53

[24] Kamijou T, Nakajima T and Ozawa H 1994 Effects ofosteocytes on osteoinduction in the autogenous rib graft inthe rat mandible Bone 15 629–37

[25] Heiple K G, Goldberg V M, Powell A E, Bos G D andZika J M 1987 Biology of cancellous bone grafts Orthop.Clin. North Am. 18 179–85

[26] Massin P, Bocquet L, Huten D, Badelon O and Duparc J 1995Observations radiographiques et histologiques d’allogreffesautoclavees et non autoclavees dans la metaphyse femoraleinferieure de chien Rev. Chir. Orthop. Reparatrice Appar.Mot. 81 189–97

[27] Dumas A, Gaudin-Audrain C, Mabilleau G, Massin P,Hubert L, Basle M F and Chappard D 2006 The influence ofprocesses for the purification of human bone allografts onthe matrix surface and cytocompatibility Biomaterials27 4204–11

6