J Clin Periodoniol 1998: 25: 721-727Printed in Denmark Ali rights reserved

Copyright © Munksgaard 1998

diitical periodontologyISSN 0303-6979

The mucosal attachment atdifferent abutmentsAn experimental study in dogs

I. Abrahamsson^ T. Berglundh'',P.-O. Glantz ,̂ and J. Lindhe^^ Department of Periodontology, GoteborgUniversity. Sweden; ^Department ofProsthodontics, Lund University, Sweden

Abrahamsson I, Berglundh T. Giant: P.-O. Lindhe J. The mucosal attaehment atdifferent abutments. An experimental study in dogs. J Clin Periodontol 1998; 25:721-727. © Munksgaard, 1998.

Abstract. The present experiment was performed to exatnine if the material usedin the abutment part of an implant system influenced the quality of the mucosalbarrier that formed following implant installation. 5 beagle dogs were includedin the study. The mandibular premolars and the 1st. 2nd and 3rd maxillarypremolars were extracted. Three fixtures of the Branemark System® were installedin each mandibular quadrant (a total of 6 fixtures per animal). Abutment connec-tion was performed after 3 months of healing. In each dog the following typesof abutments were used: 2 •'control abutments'" (c.p. titanium). 2 "ceramic abut-ments" (highly sintered AUOj), 1 "gold abutment", and 1 "short titanium abut-ment". This "short titanium abutment" was provided with an outer structuremade of dental porcelain fused to gold. Following abutment connection a plaquecontrol program was initiated and maintained for 6 months. The animals weresacrificed and perfused with a fixative. The mandibles were removed and eachimplant region was dissected, demineralized in EDTA and embedded inEPON*. Semithin sections representing the mesial, distal, buccal and lingual as-pects of the peri-implant tissues were produced and subjected to histological exam-ination. The findmgs from the analysis demonstrated that the material used inthe abutment portion of the implant infiuenced the location and the quality ofthe attachment that occurred between the periimplant mucosa and the implant.Abutments made of c.p. titanium or ceramic allowed the formation of a mucosalattachment which included one epithelial and one connective tissue portion thatwere about 2 mm and 1- 1.5 mm high, respectively. At sites where abutments madeof gold alloy or denta! porcelain were used, no proper attachment formed at theabutment level, but the soft tissue margin receded and bone resorption oc-curred. The abutment fixture junction was hereby occasionally exposed and themucosal barrier became established to the fixture portion of the implant. It wassuggested that the observed differences were the result of varying adhesive prop-erties of the materials studied or by variations m their resistance to corrosion.

Key words: abutment; biomaterials: dogs;bistometry; Implant; peri-implant mucosa

Accepted for publication 23 December 1997

The mucosa that surrounds dental im-plants made of commercially pure (c.p.)titanium has been studied in man aswell as in different animal models (Ad-ell et al. 1986. Berglundh et al. 1991.1992. 1994. Buser et al, 1992. Ericssonet al, 1992, 1995, Abrahamsson et al.1996, Berglundh & Lindhe 1996.Coehran et al. 1997). It was demon-strated that the portion of the mucosathat faces the surface of the titaniumabutment can be divided into 2 different

zones; one marginal zone that harborsa junctional epithelium and one moreapical zone that is comprised of a fiberrich connective tissue (Berglundh et al.1991. Buser et al. 1992. Abrahamssonet al. 1996). From experiments in vitro,Gould et al. (1981). and in vivo Gouldet al. (1984) concluded that the junc-tional epithelium of the peri-implantmucosa via hemidesmosomes is at-tached to the titanium surface, whileBerglundh et al. (1991, 1994) and Buser

et al. (1992) from experiments in thedog suggested that the connective tissuein the interface zone has the characterof a scar (sparse in cells and vascularstructures but rich in collagen fibers)which is firmly attached to the abut-ment. The importance of this epithehal/connective tissue attachment for themaintenance of osseointegration wasemphasized by, e.g.. Bergiundh et al.(1992), Abrahamsson et al. (1996) andBerglundh & Lindhe (1996),

722 Abrahatnsson et al.

In attempts to improve esthetics andto facilitate technical processing, ma-terials other than c.p, titanium, e.g..gold alloys, ceramics and dental por-celain have been tried in the abutmentpart of different implant systems.Knowledge, however, regarding softtissue healing and characteristics of themucosal attachment to such non-ti-tanium abutment materials is essenti-ally lacking.

The aim of the present e.xperimentwas to examine how the material usedin the abutment part of a 2-stage im-plant system influenced the compo-sition and quality of the mucosal bar-rier that formed following abutmentconnection.

Material and Methods

5 beagle dogs, about 1 year old. wereincluded in the study. At the start of theexperiment the mandibular premolars(4P4. iP-i. 2P2. |P|) and the 1st. 2nd and3rd maxillary premolars ( 'P'. -P-. 'P ' )were extracted. After 3 months of heal-ing. 3 fixtures of the Branemark Sys-tem* (Nobel Biocare .AB, Goteborg.Sweden: 7x3,75 mm) were installed ineach mandibular quadrant (a total of 6fixtures per animal). Abutment connec-tion was performed after another 3months. The abutments used and theirposition were the following (Fig.l).

A/F

Fig. I. Schematic drawing illustrating thevarious landmarks used for the histometricmeasurements. PM - the marginal portion ofthe periimplani mucosa. aJE - the level of theapical termination of the junctional epithel-ium. B - the marginal level of bone to im-plant contact. A,'F - the abutment/fixtureborderline.

• 2 (1 on each side) "control abut-ments'" (standard abutment. Brane-mark System*. Nobel Biocare AB.Goteborg, Sweden) made of com-mercially pure titanium (height 5.5mm, 0 4.5 mm).

• 2 (1 on each side) "ceramic abut-ments" (highly sintered: 99.5'>ltAI2O3): CerAdapt*' (Nobel BiocareAB, Goteborg, Sweden: height 4.0and 5.5 mm, O 4,5 mm).

• I "gold abutment" (Nobel BiocareAB. Goteborg. Sweden), the alloycontained Au: 60%), Pt: 19"/,,, Pd:20"/.i and Ir: r'-'.i (height 4.0 mm. 04.5 mm),

• 1 "short titanium abutment" (Es-theti-Cone^'^. Branemark System*.Nobel Biocare AB, Goteborg.Sweden). The "short titanium abut-ment" (height 1.0 mm. 0 4.5 mm)was provided with a super structure,made of dental porcelain fused togold, which gave an overall geometryand dimension similar to the control-. gold- and ceramic abutments(height 4.0 mm and 0 4.5 mm).

.All abutments were given final sur-face treatment by the manufacturer(Nobel Biocare AB, Goteborg.Sweden).

The abutments were tightened with aTorque Controller (Nobel Btocare AB,Goteborg. Sweden I. This device was setat 20 Ncm when the "control", the"gold" and the "short titanium" abut-ments were installed and set at 32 Nemwhen the "ceramic abutments" wereplaced.

Following abutment connection, a 6-month period of plaque control, includ-ing daily cleaning with toothbrush anddentifrice, was initiated. At the end ofthis period, a clinical examination in-cluding assessment of plaque and softtissue inflammation was performed.The animals were sacrificed with anoverdose of Sodium-Pentothal and per-fused with a fixative through the carotidarteries. The fixative consisted of a mix-ture of S".''» glutaraldehyde and A'Vaformaldehyde buffered to pH 7.2 (Kar-novsky 1965), The mandibles were re-moved and each implant region was dis-sected using a diamond saw (Exakt*,Kulzer, Germany). The tissue samples,containing the implant and the sur-rounding soft and hard peri-implanttissues, were placed in EDTA andfurther processed using a modificationof the "fracture technique" (Thoms-en & Ericson 1985) as described byBerglundh et al. (1994). Thus, before

the hard tissue was fully decalcified,cuts, parallel with the long axis of theimplants, were made through the entireperi-implant tissue. The cuts were madeat the mesial and distal aspects of theimplant and allowed a proper dissectionof the peri-implant tissues into mesio-buccal, mesio-lingual. disto-buccal anddisto-lingual units. Decalcification wascompleted in EDT.A and dehydrationperformed in serial steps of ethanoiconcentrations. Secondary fixation inOSO4 of the tissue samples was carriedout and the units were finally embeddedin EPON" (Schroeder 1969). Sectionswere produced from each tissue unitwith the microtome set at 3 /mi. Thesections were stained in PAS and tol-uidine blue (Schroeder 1969). 3 sec-tions, representing the central part ofeach of the 4 units, were selected andsubjected to histological examination,

Hisiometric analysis

In each section, histometric measure-ments were performed to describe thevertical dimension of the marginal peri-implant tissues adjacent to the implant.The following landmarks were iden-tified and used for the linear measure-ments (Fig, I); PM - the marginal posi-tion of the peri-implant mucosa. aJE -the apical termination of the jtinctionalepithelium. B - the marginal level ofbone to implant contact: .AT - the levelof the abutment/fixture border.

The distances between the landmarkswere measured m a Leica DM-RBE*microscope (Leica. Germany) equippedwith an image system Q-500 MC" (Le-ica Germany).

Morphometric analysis

The morphometric measurements wererestricted to a 80 fim wide zone of theconnective tissue lateral to the abut-ment surface and between aJE and A/F.The analysis included assessments ofthe content of collagen (Co), vessels(V). fibroblasts (Fi). leukocytes (Leu)and residual tissue (R: the remainingtissue constituents such as nerves andmatrix components lumped together)and was carried out in a Leica DM-RBE* microscope (Leica, Germany)equipped with an image system Q-500MC* (Leica, Germany), A lattice com-prising 100 light points (Schroeder &Munzel-Pedrazzoli 1973) was superim-posed over the connective tissue area ata magnification of XIOOO.

Mucosal attachment at abutments 723

Statisticai anaiysis

For each dog. mean values from meas-urements made at the mesiobuceal. dis-tobuccal, mesiolingual and distolingualunits were calculated for the various di-mensions and for the different types ofabutment. Differences between the vari-ous abutment sites were analyzed usingthe Student /-test for paired obser-vations. The null hypothesis was re-jected at /j<0.05.

ResultsClinical observations

Healing following fixture installationand subsequent abutment connectionwas uneventful. During the plaque-con-trol period a minor fracture occurred at2 of the "ceramic abutments". The frac-tures involved about 2 mm of the co-ronal part of the abutment and did notinterfere with the adjacent mucosa. Theclinical examination performed at theend of the study disclosed that all abut-ment surfaces in all five dogs were virtu-ally free from plaque and that the peri-implant mucosa at all implant sites wasdevoid of clinical signs of inflammation(Fig. 2). Marked soft tissue recession,resulting in an exposure of the abut-ment/fixture junction was observed atthe mid-buccal and.'or mid-lingual as-pects at 3 out of 5 of the "gold abut-ment" sites.

Gross histoiogicai and histometricobservations

The hislological landmarks could easilybe identified in all sections.

The results of the histometric meas-urements are reported in Table 1,

Control abutment I TitaniutitjThe peri-impiant mucosa adjacent tothe "control abutment" was comprisedof one epithelial and one connectivetissue portion. The connective tissue

portion, located between aJE and B("zone of connective tissue inte-gration"), was characterized by a densecollagenous network including few vas-cular structures and scattered fibro-blasts and inflammatory celis (Fig, 2).The mucosa was lined by a keratinizedepithelium which was continuous witha thm junctional epithelium facing theimplant. The height of the junctionalepithelium (PM - aJE) was about 2mm. The most marginal position ofbone to implant contact (B) was locatedat a distance of 0.78 mm "apical" of A/F and 1.3 mm "apical" of alE. The dis-tance between the mucosal margin(PM) and A/F was on the average 2.54mm. An inflammatory cell infiltrate wasconsistently identified in the connectivetissue at the level of A/F. This infiltratehad an "apico-coronal" dimension ofabout 0.6 mm and occupied an area ofabout 0.10 mm-.

Ceramic abutmentThe peri-implant mucosa at the "cer-amic abutment" had many features mcommon with the soft tissue at the"control abutment" sites. Thus, the mu-cosal attachment comprised a 2 mmlong junctiotial epithelium and a zoneof connective tissue that was about 1.3mm high (Fig. 3). The distance between.A/F and B was on the average 0,80 mmand the distance between PM and A/Fwas 2.56 mm. In all sections, an in-flammatory cell infiltrate was observedm the connective tissue at the level ofA/F (Fig. 3). This lesion was larger thanthe corresponding lesion identified atthe "control abutment" sites. Thus, the"apico-coronal" dimension and the sizeof the ICT were about 0.8 mm and 0.18mm-, respectively.

Gold abuttnentThe peri-implant mueosa surroundingthe "gold abutment" (Fig. 4) differed in

several aspects from the mucosa at thecontrol sites. Thus, the dimension PM -B at the "gold abutment" sites was sig-nificantly smaller than the correspond-ing dimension at the control sites: 2.55mm versus 3.32 mm. Consequently, thelength of the junctionai epithelium andthe height of the zone of connectivetissue attachment were consistentlysmaller at the "gold abutment" than atthe "control abutment" (1.75 mm ver-sus 2.04 mm and 0.79 mm versus 1.28mm respectively). B was located a dis-tance of 1.80 mm "apical" of A/F, i.e.about 1 mm further "apical" comparedto its location at the "control abut-ment". This difference was statisticallysignificant. The distance PM - AJ'F wasalso significantly smaller at the "goldabutment" (0.75 mm) than at the con-trol sites (2.54 mm).

Short titaniutn abutmentThe "short titanium abutment" wassurrounded by a mucosa the height ofwhich was smaller than the correspond-ing soft tissue at the control sites: 2.99mm versus 3.32 mm. Also the length ofthe junctiona! epithelium and the heightof the connective tissue attachmentwere somewhat smaller than the corre-sponding dimensions at the controlsites: 1.81 mm versus 2.04 mm and 1.19mm versus 1.28 mm. respectively. Land-mark B was located a distance of 1.26mm "apical" of A/F, i.e. about 0.5 mmmore "apical" than at the control sites.This difference was statistically signifi-cant. The distance P - A/F was at the"short titanium abutment" sites sig-nificantly smaller than at the controlsites (1.73 mm and 2.54 mm, respec-tively). The connective tissue compart-ment lateral to A/F and to the junc-tional epithelium (Fig. 5), consistentlyharbored an inflammatory cel! infil-trate. The "apico-coronal" dimensionof this lesion was about 0.5 mm and thesize was about 0.07 mm-.

Table I. 'Vertical dimensions (mm) of the peri-implant soft tissue component

(mm)

controlceramicgoldshort titanium

PM-B

mean

3.323.362.55*2.99

SD

0.240.260.28

PM-aJE

mean

2.042,011,751,81

SD

0,220.440.200.36

aJE-B

mean

1.281.340.791.19

SD

0.1!

0.31*0.38

PM-A

mean

2,542,560.75*1.73*

/F

SD

0.350.220.440.43

A,

mean

0.780.801.80*1.26*

F-B

SD

0,170,160,210,31

Results from the hisiometric measurements: PM: the marginal portion of the peri-implant mucosa. aJE: the level of the apical termination ofthe junctional epithelium. A/F: the abutment/fixture borderline. B: the marginal level of bone to implant contact. Mean values and standarddeviations (SD). * Indicates statistically significant difference versus Control (titanium). /'-value< O.O.'i.

724 Abrahamsson et al

Mueosa! attachment at abutments 725

Table 2. Results from the morphometricmeasurements.

(%)

CoVFiLeuR

Control

mean

84.04.07.21.33.5

SD

2.90.11.51.00.5

Ceramic

mean

S2.24.37.32.33.9

SD

4,41,21.01.60.7

The volume Va of the connective tissue occu-pied by collagen (Co), vascular structures(V). fibroblasts (Fi), leukocytes (Leu) and re-sidual tissue (R). Mean values and standarddeviations (SD).

Morphometric observations

The morphometric analysis of the peri-implant mucosa was restricted to the"control abutment" and the "ceramicabutment" sites. At the "gold abut-ment" and the "short titanium abut-ment" sites, the mucosal margin duringhealing consistently receded. Hence, theconnective tissue component of the mu-cosal attachment became located at thefixture which at all experimental siteswas made of c.p. titanium.

The volume fraction of the connec-tive tissue occupied by collagen, variedbetween 84 '>!. and 82 "/a at the "controlabutment" and the "ceramic abutment"sites (Table 2), Fibroblasts made upabout 7 "/a and vascular structuresabout 4 "Al. Only a few scattered m-flammatory cells could be identified inthe zone of connective tissue attach-ment.

Discussion

The present experiment demonstratedthat the material used in the abutment

portion of the implant was of decisiveimportance for the quality of theattachment that occurs between the mu-cosa and the implant. Hence, abutmentsmade of c.p, titanium or highly sinteredaluminum based ceramic (AKO, ) (l) es-tablished similar conditions for mucosalhealing to the abutment surface and (ii)allowed the formation of an attachmentthat mcluded one epithelial and oneconnective tissue portion that wereabout 2 mm and 1-1.5 mm high.

On the contrary, at sites where abut-ments made of gold alloy or dental por-celain were installed at the second stagesurgery, no proper attachment seemedto form at the abutment level but thesoft tissue margin receded and bone re-sorption occurred. The abutment fix-ture junction was hereby occasionallyexposed and the mucosal "seal" was es-tablished to the fixture portion of theimplant.

The finding that the mucosal attach-ment that formed to an abutment madeof c,p. titanium was comprised of onezone of epithelium (about 2 mm) andone zone of collagen rich/cell poor con-nective tissue (about l-i.5 mm) is con-sistent with data previously reportedfrom similar experiments in the dog(Berglundh et al. 1991. 1992. Ericssonet al. 1996. Abrahamsson et al. 1996.Cochran et al. 1997). The current find-ings also confirm data by Berglundh &Lindhe (1996) who in the beagle dogmodel studied the soft tissue attach-ment that formed at sites with a thin(about 2 mm) mucosa. The authors re-ported that during healing followingabutment connection at such sites boneresorption occurred (about 1.5 mm) toallow the formation of a mueosal bar-rier that included both epithelium (2mm) and connective tissue (1.3 mm).

The observation that the "ceramicabutment" established healing con-ditions similar to those at the controlabutment confirms findings previouslyreported from animal experiments andciinicai trials using so called single crys-tal sapphire implants (McKinney et al.1985. Hashimoto et al. 1988, 1989. Ak-agawa et al. 1989. Fartash et al. 1990.Arvidsson et al. 1991). It was suggestedfrom both light- and electron micro-scopic examinations that the junetionalepithelium that forms at the abutmentpart of such AKO^ based implants hasmany features m common with thedento-gingival epithelium at teeth(McKinney et al, 1985. Fartash et al.1990) and that the AFOj material wasinert (i.e. showed no release of alumi-num in e.g. aggressive acidic test solu-tions) and biocompatible (i.e. showedno cell toxicity. minimal foreign bodyreaction and allowed ceil proliferationon its surface) (Arvidsson et al. 1991)..Arvidsson et al. (1996) compared theperiimplant mucosa at Branemark*type implants and the mucosa thatformed at single crystal sapphire AKOiimplants. Soft tissue biopsies were re-trieved from human volunteers andexamined by histological means. Themucosal units that were present at the 2implant systems had a similar compo-sition regarding both epithelial andvarious connective tissue elements. Thisfinding is supported by observationsmade in the current study.

An important observation made inthe present experiment relates to the ap-parent inability of the gold alloy to pro-mote a mucosal healing that includes azone of connective tissue attachment or"connective tissue integration" (Berg-lundh et al. 199!). On the contrary, dur-ing healing the mucosal margin at the

Fig. 2. Bucco-lingual cross section of the peri-implant tissues at the level of the abutment/fixture borderline of a "control abutment" site. Thehistological landmarks are depicted in Fig, I, P.A.S and toluidine blue, original magnification XIOO.

Fig. 3. Bucco-lingual cross section of the peri-implant tissues at the level of the abutment/fixture borderline of a "ceramic abutmeni" site. Notethe infiammatory infiltrate in the connective tissue lateral to the abutment ' fixture borderline (small arrows). The histological landmarks aredepicted in Fig. 1. PAS and toiuidine blue, original magnification xlOO,

Fig. 4. Bucco-hngual cross section of the marginal peri-implant tissues at a "gold abutmeni" site. Note the marked recession of the peri-implant soft and hard tissues. The histological landmarks are depicted in Fig, I, PAS and toluidine blue, original magnification x50.

Fig. 5. (a) Bucco-lingual cross section of the marginal peri-implant tissues ai a "short titanium abutment". The histological landmarks aredepicted in Fig. 1. Additional landmark: C/A - the borderline between the "short titanium abutment" and the crown made of porcelain fusedto gold. Note that aJE is located apical to .A.'F. PAS and toluidine blue, original magnification X50. (b) Higher magnification of the peri-implant tissues adjacent to the most coronal bone to implant contact of the implant site illustrated in (a). The histologicaJ landmark is depictedin Fig, I, PAS and toluidine blue, original magnification XIOO. (c) Higher magnification of the tissue compartment lateral to A/F (from a)).Note the inflammatory infiltrale (small arrows) lateral to A F The hisiological landmarks are depicted in Fig. 1. PAS and toluidine blue,original magnification X200.

. 726 Abrahamsson et al.

gold abutment sites consistently receded"apically" and was at the end of thestudy in 3 otit of 5 cases found on thefixture part of the implant. At these 3abutment sites, the entire mucosalattachment (i) occurred to the titaniumsurface of the fixture, and (ii) includedone zone of epithelium and one zone ofconnective tissue. At the 2 remainingnon-exposed gold abutment sites, apocket epithelium made up the mucosallining at the abutment level, while themucosal attachment was establishedapical of the abutment/ fixture junction.In order to allow for the formation of amucosal barrier (i.e. zone of junctionalepithelium and connective tissue) atsuch sites, substantial resorption ofmarginal bone tissue occurred. This ob-servation is an agreement with Ber-glundh & Lindhe (1996) who concludedfrom an experiment in the dog that aminimum amount of soft tissue attaeh-ment (epithelium and connective tissuelto titanium is required and if not avail-able, marginal bone resorption will oc-cur to accommodate this attaehment.

The validity of this concept is sup-ported by findings made at the "shorttitanium abutment" sites. At such sitesthe available titanium surface of theabutment was only 1 mm. and. hence,insufficient for a properly dimensionedm.ucosal attachment. As a result, sig-nificant amounts of bone resorption oc-curred at such sites during healing. Inthis context, however, it should be ob-served that in the present experimentthe degree of bone resorption was morepronounced at the gold abutment thanat the "short titanium abutment" sites(1,80 mm vs. 1.26 mm).

The finding that the gold alloy abut-ments failed to properly interact withthe connective tissue is in agreementwith findings recently reported in astudy Ln rabbits by Thomsen et al.(1997). They studied the interface be-tween cortical bone and implants madeof gold, zirconium and titanium and re-ported that the amount of bone formedwithin the threads of the screw type fix-ture tested, as well as the degree of boneto implant contact was smaller at goldthan at titanium and zirconium im-plants. Furthermore, areas of soft con-nective tissue including multinuelearcells and macrophages were observedmore frequently at gold implants thanat implants made of titanium and zir-conium.

It is suggested that the differences re-ported by Thomsen et al. (1997), and

also observed in the present experiment,regarding the tissue interaction with thevarious implant surfaces may be due todivergence in the basic bio-adhesiveproperties for the materials studied.Thus, the bio-adhesive properties areknown to be comparatively poor formetals but more optimal for ceramicsand ceramic like materials (e.g. titaniumdioxide) (for review, see Meyer et al.(1991). It is furthermore commonly ac-cepted that ceramics and titanium diox-ide are more resistant to corrosion thangold alloys. In other words, the surfacelayers of titanium and ceramics arechemically more stable (Handbook ofChemistry and Physics) and thereforewill allow cells to grow in contact withthe surface.

Finally, in all specimens representingthe "control abutment" and the "'cer-amic abutment" sites, infiltrates of in-flammatory cells were consistently pres-ent in the mucosa lateral to the abut-ment/fixture junction. This so called"abutment ICT" was previously de-scribed by, e.g., Ericsson et al. (1995)and Abrahamsson et al. (1998) as thehost response to bacterial contami-nation of the internal portion of theBranemark System* of dental implants.In the present experiment the "abut-ment ICT" was somewhat larger in theceramic abutment sites than in the con-trols. It is likely that this difference canbe explained by the presence of a largermicrogap between the ceramic abut-ment and the fixture than between thetitanium abutment and the fixture atthe control site.

Acknowledgments

This study was supported by grantsfrom Nobel Biocare AB. Goteborg.Sweden and Colgate Palmolive Co, Pis-catawav. NJ, USA.

Zusammenfassung

Das Attachmeni der Muko.ia bei ver.schiede-nen Imptantatpfosten. Eine experimentelleStudie an HundenDas hier beschriebene Experiment wurdevorgenommen um zu untersuchen, ob derWerkstoff eines Implantatpfostens die Quali-tat der Mukosabarriere beeinfluGt. die sichnach dem Einbringen des Implantats bildet.5 Beagle Hunde wurden in die Studie einbe-zogen. Die Unterkieferpramolaren und derL. 2. und 3. Oberkieferpramolar wurden ex-trahiert. 3 Fixturen des Branemark System®wurden in jeden Unterkieferquadranten ein-gebracht (insgesamt 6 Fixturen pro 'Versuchs-

tier). Nach 3 Monate langer Einheilung wur-de die Pfoslenmontage vorgenommen, Bei je-dem Hund kamen folgende Pfostentypen zurAnwendung: 2 "Kontrollpfosten" (c,p. Ti-tan). 2 "keramische Pfosten" (hochgesinter-tes All Oj). 1 Pfosien mit Gold als Werkstoffund ein "kurzer Titanpfosten", Der "kurzeTitanpfosten" wurde mit einem iiuBeren Be-lag von Gold-Aufbrennkeramik versehen.Nach der PI'ostenmontage wurde ein Plaque-kontrollprogramm eingeleilet und 6 Monatelang beibehalten. Die Versuchstiere wurde ge-opfert und mitlels Perfusion fixiert. Die Un-terkiefer wurden entfernt und jede Implanta-tregion freipritpariert. in EDTA deminerali-siert und in EPON* eingebettet, HalbdunneSchnittpraparate. die die periimplantarenGewebe in mesialer. distaler. bukkaler undlingualer Sicht darsteliten. wurden angefcr-tigt und histologisch untersucht. Die Ergeb-nisse der Analyse zeigten. daC der Werksloffdes Implantatpfostens Position und Qualiliitdes Attachments zwischen periimplanliirerMukosa und Implantat beeinfluflte. lmplan-tatpfosten. die aus c,p. Titan oder Keramikhergestelll waren. ermoglichten die Bildungeines mukosaien Attachments mit einer etwa2 mm bzw: 1-1.5 mm hohen epithelialen undbindegeweblichen Manschette, .An Stellen milImplantatpfosten aus GoSdIegierung oderdentaler Keramik (dental porcelain) bildetesich kein richtiges Attachment in Pfostenho-he, Der Rand des Weichgewebes zog sich zu-rlick und es kam zu Knochenresorption, DerUbergang von der Fixtur zum Pfosten wurdegelegendlicb i'reigelegt und die Mukosabar-riere etablierte sich an dem E'ixturenteil desImplantats, Es erscheint wahrscheinlich. daBdie beobachleten Unterschiede als das Resul-tat verschiedenartiger adhasiver Eigenschaf-ten des Werkstoffs oder als Variationen un-terschiedlicher Korrosionsresistenz aufgefaCtwerden miissen.

Resume

L 'attache muqueu.w au niveau de differents pi-liers Une etude experitnentale .sur Ic chienL'etude presente a eie realisee pour examinersi le materiel utilise dans la partie pilier d'unsysteme implantaire influen<;ait la qualite dela barriere muqueuse qui se formait apresrinsiallalion de I'implant, 5 chiens Beagleont ete inclus dans cette etude, Les prenrolai-res mandibuiaires ei les premieres, deuxiemeset troisiemes premolaires maxillaires ont eteavulsees. 3 implants ad modum Branemarkont ete places dans chaque quadrant mandi-bulaire totalisant 6 implants par animal. Laconnexion du pilier a ete effectuee apres 3mois de guerison, Chez chaque chien les ty-pes suivants de piliers ont ete utilises: 2 pi-liers contr6les (titane commercialemenl pur).2 piliers en ceramlque (99.5'yit AI2O3). un pi-lier en or et un pilier en titane court (structu-re externe faite en porcelaine dentaire sur or).Apres la connexion du pilier un programmede contr6le de la plaque dentaire a demarreet a ete maintenu durant six mois. Les ani-maux ont ele lues, les mandlbules otees et

Mucosal attachment al abutments 727

chaque region implanlaire a ete dissequee.demineralisee dans 1' EDTA e! indue dansI 'EPON'. Des coupes semi-fines represen-tant les aspects mesiaux. distaux. vestibuiai-res et linguaux des tissus paroi'mplantairesont ele produiles el soumises a I'examen his-lologique, Les decouvertes de cette analyseont demonlre que ie materiel utilise dans laportion pilier de I'implanl influenfail la loca-lisation et la qualite de I'atlache qui se faisaitentre la muqueuse paro'i'mplantaire et I'im-plant, Les piliers faits en titane commerciale-ment pur ou en ceramique donnaient nais-sance a la formation d'une attache muqueusequi comportait une partie epitheliale et unepartie de tissu conjonctil'qui etaient respecti-vement d'environ 2 mm el de 1-1,5 mm dehauteur, Au niveau des site.s ou les piliersetaient i'abriques en or ou en porceiaine au-cune attache correcle ne se formail au niveaudu pilier mais ie lissu mou marginal availsubi une recession et s'accompagnait duneresorption osseuse. La jonction implant-pi-lier etait occasionnellement mise a nu et labarriere muqueuse etait elablie sur la portionde rimplant lui-meme, Les differences obser-vees seraient le resullal de proprietes d'adhe-sion differenles des materiaux etudies oudues aux variations dans leur resistance a lacorrosion.

References

.Abrahamsson. I,. Berglundh. T . Wennslrom.J. & Lindhe. J, (1996) The peri-implanthard and soft tissues at different implantsystems, A comparative study in the dog.Clinical Oral Implants Research 7. 212-219.

.Abrahamsson. I.. Berglundh. T. & Lindhe. J.(1998) Soft tissue response to plaque-for-mation at different implant systems, .Acomparative study in the dog. ClinicalOral Implants Researeh. 9. 73-79.

Adell. R., Lekholm. U . Rockier. B.. Brane-mark. P-E. Lindhe. J.. Eriksson. B, &Sbordone. L, (1986) Marginal tissue reac-tions al osseointegraled titanium fixtures(II, A 3 year longitudinal prospectivestudy. International Journal of Oral &Maxillofacial Surgery 15. 39-52.

Akagawa. Y.. Takata. T . Matsumoto. T . Ni-kai. H. & Tsuru. H, (1989) Correlation be-tween clinical and histological evaluationsof the peri-implant gingiva around thesingle-crystal sapphire endosseous im-plant. Journal of Oral Rehabilitation 16.581-587,

Arvidsson. K.. Fartash. B.. Moberg. L-E..Grafstrom. R, & Ericsson. 1, (1991) In vi-tro and in vivo experimental studies onsingle crystal sapphire dental implants.Clinical Oral Implants Research 2, 47-55.

Arvidsson. K.. Fartash, B., Hiiliges. M, &Kondell. PA, (1996) Histological charac-teristics of peri-implant mucosa aroundBranemark and single-crystal sapphire im-plants. Clinical Oral Implants Re.seurch 7.1-10.

Berglundh. T . Lindhe. J.. Ericsson. L. Mar-inello. C.P. Liljenberg, B, & Thomsen. R(1991) The soft tissue barrier al implantsand teeth. Clinical Oral Implants Research2, 81-90,

Berglundh. T. Lindhe. J.. Marinello. C , Er-icsson. 1, & Liljenberg. B. (1992) Softtissue reaction to de novo plaque forma-tion on implants and teeth. Clinical OralItnplatits Research 3. 1-8,

Berglundh. T.. Lindhe. J.. Jonsson. K, & Er-icsson. 1. (!994) The topography of thevascular systems in the periodontal andperi-implani tissues in the dog. Journal ofClinical Periodontulogy 21. 189-193.

Berglundh, T, & Lindhe. J. (1996) Dimensionof the peri-implant mucosa. Biologicalwidth revisited. Journal of Clinical Period-ontology 2i. 91\-913.

Buser. D . Weber. H.R. Donath. K.. Fiorelli-ni, JR . Paquette. D.W & Wilhams. R,C,(1992) Soft tissue reactions to non-sub-merged unloaded titanium implants inbeagle dogs, Jourtiat of Periodontology 63.226-236.

Cochran. DL. Hermann. J.S,. Schenk. R.K..Higginbouom. FL. & Buser. D. (1997)Biologic width around titanium implants.A histometric analysis of the implanto-gingival junction around unloaded andloaded nonsubmerged implants in the can-ine mandible. Journal of Periodontology68. 186-198.

Ericsson. 1. Berglundh. T. Marinello. C.P..Liljenberg. B, & Lindhe. J, (1992) Long-standing plaque and ginghitis at implantsand teeth in the dog, Ciinicai Orai Im-plants Research 3. 99-103,

Ericsson. L. Persson. L.G,. Berglundh. T .Marinello. C.P. Lindhe. J. & Klinge. B,(1995) Different types of inflammatory re-actions in peri-implant soft tissues. Journalof Clinical Periodontology 22. 255-26!,

Ericsson. L. Nilner. K.. Klinge. B, & Glantz.P-O. (1996) Radiographical and histologi-cal characteristics of submerged and non-submerged titanium implants. ClinicalOral Implants Research 7. 20-26,

Fartash. B.. .Arvidsson. K. & Ericsson. 1.(1990) Histology of tissues surroundingsingle crystal sapphire endosseous dentalimplants. Clinical Oral Implants Re.seareh1. 13-21.

Gould. T.R.L,. Brunetle. DM, & 'Westbury.L. (1981) The attachmeni mechanism ofepithelial cells to litanium in vitro. Journalof Periodonial Re.warch 16. 611-616.

Gould. T.R.L.. Westbury. L. & Brunelle.

DM. (1984) Ultrastructural study of theattachment of human gingiva to titaniumin vivo. Journal of Prosthetic Dentistry 52.418^20.

Handbook of Chemistry and Physics(1971).The Chemical Rubber Co, Cleve-land. Ohio,

Hashimoto. M,. Akagawa. Y.. Nikai. H. &Tsuru. H (1988) Single-crystal sapphireendosseous dental implant loaded withfunctional stress - clinical and histologicalevaluation of peri-implam tissues. Journalol Oral Rehabilitation 15. 65-76,

Hashimoto. M,. Akagawa. Y.. Nikai. H, &Tsuru. H (1989) Ultraslructure of the per-implant junctional epithelium on single-crystal sapphire endosseous dental im-plant loaded with functional stress.Journal of Oral Rehabilitation 16. 261-270.

Karnovsky. M. J, (1965) A formaldehyde-glutaraldehyde fixative of high osmolarityfor use in electron microscopy. Journal ofCell Biology 27. 13 7.A-13 8 A.

McKinney. R,V.. Stefiik. D F , & Koth, D,L,(1985) Evidence for a junctional epithelialattachment to ceramic dental implants.Journal of Periodontology 56. 579-591,

Meyer. .A.E.. Baier. RE. . Glanlz. P-O. & Na-tiella. J.R. (1991) Biomaterials: Selection.evaluation, and preparation. In: Babbush.C.A. (ed,). Dental Implants. Principles andpractice, pp, 31^2 . WB. Saunders Comp..Philadelphia,

Schroeder. H.E, (1969) Ultrastructure of thejunctional epithelium of the human gin-giva. .Acta Helvetica Odontologiea 13. 65-83.

Schroeder. H.E, & Munzel-Pedrazzoli. S.(1973) Correlated morphometric and bio-chemical analysis of gingival tissue.Journal of Microscopy 99. 301-329.

Thomsen. P. Ericson, L.F, (1985) Light andtransmission electron microscopy used lostudy the tissue morphology close to im-plants, Biomaterials 6. 42 i^24 .

Thomsen. P. Larsson. C . Ericson. L.E,. Sen-nerby. L.. Lausmaa. J, & Kasemo. B,(1997) Structure of the interface betweenrabbit cortical bone and implants of gold.zirconium and titanium. Journal of Ma-terials Science: Materials in Medicine 8.653-665.

.Address:

Ingemar AbrahamssonDepartment of Periodontology.Institute of OdontologyGoteborg UniversityBox 450SE 405 30 GoteborgSweden