The soft tissue barrier at implants and teeth
Berglundh T. Lindhe J, Eriesson 1, Marinetlo CP. Liljenberg B. Thom-
sen P. The soft tissue barrier at implants and teeth.
Clin Oral Impl Res 1991: 2: 81-99.
Abstract: In the present animal experiment. analyses and comparisons
were made between the structure and composition of clinically healthy
supraalveolar soft tissues adjacent to implants and teeth. 5 beagle dogs
were used. The right mandibular premolar region was selected in each
dog for placement of titanium implants, while the left mandibular pre-
molar region served as control. Extractions of the mandibular premolars
were performed. healing allowed, following which titanium fixtures were
installed in the edentolous premolar region, Abutment connection was
carried out 3 months later. After another 2 months of healing. plaque
control was initiated and maintained for 8 weeks. At the end of the
plaque control period. clinical examinations were performed and biop-
sies harvested from the implant site and the contralateral premolar tooth
region. Following fixation and decalcification. all tissue samples were
embedded in EPON and examined by histometric and morphometric
means. The result from the analyses demonstrated that the periimplant
mucosa which formed at titanium implants following abutment connec-
tion had many features in common with gingival tissue at teeth, Thus.
like the gingiva. the peri-implant mucosa established a cuff-like barrier
which adhered to the surface of the titanium abutment. Further. both
the gingiva and the peri-implant mucosa had a well-keratinized oral
epithelium which was continuous with a junctional epithelium that faced
the enamel or the titanium surface. In the periimplant mucosa. the col-
lagen fibers appeared to commence at the marginal bone and were par-
allel with the abutment surface. All gingival and periimplant units exam:
ined were free from infiltrates of inflammatory cells. It was suggested
that under the conditions of study. both types of soft tissues. gingi
periimplant mucosa. have a proper potential to prevent subgingival
plaque formation.
T.Berglundh,' J. Lindhe |. Ericsson,*
.P.Marinello,” B.Liljenberg’ and
P.Thomsen®
"Departnent of Priadontlony Faculty of
(Odontology, University of Gothenburg,
Gotenbug, Sweden; “Department of Fixed and
Fenovable Proshodontis,Faculy of Odontology
University of rch 2arih, Switzerland,
Deparment of Anatomy, University of
Gotenburg, Sweden
Koy words: anu - sof tssue—toth- gingiva
= hitamatie measurements = morphometric
‘Tor Berlundh, Department of Periodontology
Schoo of Dents, University of Gothenburg. Box
3907, $-40038 Gothenburg,
‘Sweden
‘Accaptd for publication 7 February 1981
The free marginal/interproximal gingiva has a bar-
rier function and guards the periodontium proper
(root cementum with investing principal fibers, pe-
riodontal ligament and alveolar bone proper) from
insults from the external environment. The protec-
tive mechanism includes both structural compo-
nents, such as (i) the junctional epithelium and its,
constituents. (ii) the valeular plexus, (iii) the colla-
gen fibers and (iv) other connective tissue ele-
ments and inflammatory reactions (vascular and/or
cellular) in sub- and intra-epithelial cell infiltrates,
In subjects who are resistant to periodontal disease
(Lée et al. 1986. Papapanou et al. 1991) this bar-
rier is obviously effective throughout life. In other
subjects. however, the protective mechanism ap-
pears inadequate. and as subgingival plaque forms
a process of periodontal tissue destruction is initi-
ated and without treatment maintained until the
teeth are lost (Van der Velden et al. 1985, Lée
et al. 1986, Sjddin et al. 1989). In this context, the
apparent longevity of threaded, commercially pure
titanium implants for the treatment of edentulous-
ness and cranio-facial defects (Branemark et al
1977) is of interest. Obviously it is through the
process of osseointegration that the stability of the
anchor part, the fixture. of the implant is secured.
The maintenance of stability and function of the
load carrying implant, however. is dependent on a
well functioning barrier mechanism established at,
the transmucosal passage of the implant (Ten Cate
1985. McKinney et al, 1988, Akagawa ct al. 1989).
Results from different investigations have empha-
‘SoeoRnEONNESENDEISEI tooth/gin-
and the implant/mucosa interface (Schroeder
etal. 1981. Jansen 1985. Gould 1985. McKinney
et al. 1988. Van Drie et al. 1988. Akagawa et al
1989, Buser ct al. 1989, Carmichael 1989). ‘#869
ing both epithelial structures and connective tissue
¢lemeEntsH The lack of root cementum on the im-
plant surface, however. also establishes certain
fundamental differences between the two unitsBerglundh et al.
Fig. 1. Outline of the study. $ beagle dogs were used, The Ist
2nd, 3rd and 4th mandibular premolars (right side) were ex:
acted, After a healing period of 3 months, the titanium
fixtures were installed and after another healing period of 3
‘months the abutment connection was performed. Abutment
connection was followed first by a healing period of 2 months
and consecutively by a plaque control period of 2 months. At
the end of the plague control period a clinical examination was
performed and biopsies harvested.
both with respect to fiber orientation and fiber
attachment (Buser et al. 1989).
The present report is the first in a series of
investigations aimed at analyzing the interface and
interaction between titanium implants and the ad-
jacent soft and hard tissues in health and e
‘The objective of the animal experiment reported
below was to compare the structure and compo:
tion of the clinically healthy supraalveolar soft tis-
sue adjacent to titanium implants with the corr
sponding gingival tissue adjacent to teeth.
Materi
5 beagle dogs were used. They were at the start of
the experiment 15 months oid. In each dog, the
right mandibular premolar region was selected for
placement of titanium implants (test) while the left
mandibular premolar region served as control
‘Thus, extractions of the mandibular right premo-
lars (4P, 3P, 2P, 1P) were performed and healing,
allowed for 3 months (Fig. 1). 3 titanium fixtures
(Branemark®; length: 7 mm, diameter: 3.75 mm,
Nobelpharma, Gothenburg, Sweden) were in-
stalled (Branemark et al. 1977, Adell et al. 1985)
in the edentulous mandibular premolar region. 3,
months following fixture installation, abutment
connection was performed (Brinemark et al
1977, Adell et al. 1985). After a 2-month healing
period, a plaque control program was initiated and
maintained for 8 weeks. During this period the
dogs were. once daily. subjected to meticulous
tooth- and titanium abutment cleaning with the use
of toothbrush and dentifrice
At the end of the plaque control period, clinical
examinations including assessment of plaque and
gingival inflammation were performed at the buc-
cal surfaces of the titanium abutments (test) and
the corresponding contralateral mandibular pre-
molars (control) (Fig. 2)
Block-biopsies were harvested from the implant
site of 2P (i.c., the region of the right mandibular
‘and methods
Fig. 2. Clinical photographs of the lower mandibular premolar
regions of one of the dogs. The photographs were obtained at
the end of the plague control period. (a) Implant region 4P. 3?
and 2P. (b) Tooth region P2, P3 and Pd
2nd premolar) (Fig. 3). The biopsies from the test,
sites included not only the entire imlant but also
the adjacent bone and the supraalveolar soft tis-
sue. Following fixation and decalcification (Ber-
glundh et al. 1990) all tissue samples were embed-
ded in EPON (Schroeder 1969).
Fig. 3. Photograph of a block biopsy harvested from the im=
plant region 2P. The biopsy included the entire implant and the
adjacent hard and soft tissues.Fig. 4. Photographs of the
implant (a) and the
EPON-embedded soft and
hard tissue (h) after
separation by the “fracture
technique”. The 2 vertical
sgrooves (arrows) in the
Implant surface (a1)
represents the cuts through
the plastic material made
to facilitate “fracture
Fig. 5. (a) Survey scanning electron micrograph of the surface
‘of the plastic-embedded tissue after separation from the im
plant. The smooth. upper surface (A) represents the tissue
Which before separation faced the coronal portion of the fix-
ture. The central and lower part of the micrograph (T) repre-
sents the tissue within the threaded part of the fixture. Bar
1000 urn. (b) Scanaing electron micrograph of an area of the
abutment surface after separation of the plastic-embedded tis-
sue. The striated topogeaphy of the surface is the result of the
machining procedure. No large plastic remnants are observed,
Bar = 10 am.
Soft tissue at implants and teeth
The EPON cylinders which included the implant
with adjacent soft and hard tissues were each di-
vided into 1 marginal and 1 apical portion. The
smaller apical portion was discarded. In the mar-
ginal portion 4 grooves were prepared in “apico-
coronal” direction. The grooves, which extended
from the periphery of the specimen to the surface
of the titanium implant, divided the specimen into
4 units (buccal, mesial. distal and lingual) out of
which only the buccal was further processed (“frac
ture technique”, Thomsen & Ericson 1985)
(Fig. 4). In order to facilitate analysis of the in-
terface tissue, i.c., the soft and hard tissue immedi-
ately adjacent to the implant. the inner surface of
each unit was sputtered with goold and the tissue
section re-embedded in EPON (Thomsen & Eric-
son 1985). The titanium abutment- and the tita-
nium fixture-surface as well as the tissue surface
were sputtered with gold and examined by scan-
ning electron microscopy (SEM) (JEOL. T-300)
(Fig. 5). Bucco-lingual sections were cut with the
microtome set at 3 wm. The sections were stained
in PAS and toluidine blue (Schroeder 1969). From
each unit 5 sections, representing the central part
of the tissue, were used for histometric and mor-
phometric analyses.
The control biopsies. ic... from the left 2nd man-
dibular premolar region, were obtained according
to a procedure described by Schroeder et al
(973)
Histometric analysis
‘The following linear distances were measured.
(a) In the sections from the implant sites
(Fig, 6b): (i) the margin of the soft tissue adjacent
to the titanium abutment (GM) to the apical termi-
nation of the junctional epithelium (aJE); (ii) the
apical termination of the junctional epithelium
(JE) to the marginal bone crest (BC)Berglundh et al.
|
—f
Fig. 6. Schematic drawing demonstrating the landmarks which were used for the assessment of certain linear distances (histometvic
measurements. (a) At the tooth section (let). (i) The gingival margin (GM) to the apical termination of the junctional epithelium
(aE). (i) The apical termination of the junctional epithelium (aJE) to the marginal bone crest (BC). (ii) The gingival margin
(GM) to the marginal bone crest (BC). (b) At the implant section (right). (i) The margin of the soft tissue adjacent to the titanium
abutment (GM) to the apical termination of the junctional epithelium (aJE). (ii) The apical termination of the junctional
epithelium (aJE) to the marginal bone crest (BC). (ii) The margin of the soft tissue adjacent to the titanium abutment (GM) to the
‘marginal bone crest (BC).
Fig. 7. Schematic drawing
itsstrating the
compartments exposed to
morphometric
measurements. At level
IV, a determination was
made of the connective
tissue proportions that
were ovcupied by
collagen (Co). vascular
structures (V),
fibroblasts (Fi).
leukocytes (Leu) and
residual tissue (R) in one
area located immediately
below the junctional
epithelium (Area 1) and in
asecond area adjacent to
the interface between the
connective tssuetitanium
implant and root
ct b cementum (Atea 1),Fig, 8. Cross-section of the buccal dento-gingival region repre:
senting 2 months of plaque control, PAS and toluidine blue
Original magnification > 40.
(b) In the sections from the tooth sites (Fig. 6a):
(i) the gingival margin (GM) to the apical termin:
tion of the junctional epithelium (aJE); (ii) the
apical termination of the junctional epithelium
{aJE) to the marginal bone crest (BC).
Morphometric analysis
‘The composition of the connective tissue with re-
spect to content of collagen (Co). vascular struc-
tures (V). fibroblasts (Fi). leukocytes (Leu) and
residual tissue (R; matrix. nerves, unidentified
cells and other tissue constituents) was determined
using a point counting procedure described by
Schroeder & Miinzel-Pedrazzoli (1973). The a
sessment were made at level IV (Schroeder &
Minzel-Pedrazzoli 1973) utilizing a Wild micro-
scope, mag. 1000 and a lattice containing 100
light points,
‘The following regions were analyzed: (i) a
100-m wide zone of the connective tissue immedi
ately lateral to the junctional epithelium (Area 1)
(Fig. 7); (ii) a 100-um wide zone of the connective
Soft tissue at implants and teeth
Fig. 9. Cross-section of the buccal and coronal part of the
periimplant hone and mucosa representing 2 months of plague
control. PAS and toluidine blue. Original magnification 40,
tissue immediately lateral to the titanium implant
or the root cementum (Area II) (Fig. 7)
Differences were analyzed using the Student
test for paired observations (N= 5),
Results
Clinical features
‘The clinical examination performed prior to biopsy
revealed that all titanium abutments as well as all
mandibular premolar surfaces were free from
plaque deposits (Fig. 2). The adjacent soft tissues.
the periimplant mucosa and the gingiva. showed
no signs of inflammation
‘The soft tissue adjacent the titanium implants
(Fig. 2a) had pink color and appeared to be firmly
adapted to the abutment surface. The borderline
between the alveolar mucosa and the attached gin
gival tissue was at the buccal aspect located about,
1-2 mm apical to the soft tissue margin. the lingual
side the corresponding distance varied between
2-4 mm.
The free and attached ging
face of the premolars (
a at the buccal sur-
ig. 2b) had pink color and
85Berglundh et al.
Tal 1 Results (mm) from the histometric measurements; mean value (x)
and standard deviation (6.0)
Test Conta! bit. Te
x 8D x SD x 80 SE
Male 24 047 205 005 009 0a ot
EBC 185 023112 003058 «028
GM-eC 380 085 317 oof 083062 at
Te various distances measured are described in Fig.6. The ference
between the implant (Test and the tooth control; mean vale standard
deviation (§.D) and standard eror (SE) Difference within 95% confidence
fr
was at the muco-gingival junction separated from
the movable alveolar mucosa. A gingival groove.
located about 1 mm apical to the gingival margin,
separated the attached gingiva from the free mar-
ginal gingiva
Gross histological appearance
The gross morphology of the soft tissue specimens,
from teeth (gingiva) (Fig. 8) and from implant sites
(Fig. 9), had many features in common. Both cate.
gories of tissue had (i) a triangular configuraiton,
(ii) a well-keratinized oral epithelium which termi-
nated at the crest of the gingival margin where
(iii) it_was continuous with a junctional epithe-
lium. In sections representing the tooth sites, the
junctional epithelium terminated at the cemento-
enamel junction (CEJ). At the implant sites, the
junctional epithelium ended at a varying distance
from the gingival margin consistently leaving
connective tissue portion coronal to the bone crest,
in direct contact with the titanium surface. ‘The
connective tissue was in both types of specimens
devoid of inflammatory cell infiltrates and charac-
terized by its collagen fiber structures.
Table 2% volume ofthe connective tissue components in the ares located
immediately below the junctional epithelium; collagen (Co), vascular struc-
tures (V}, iorobasts (Fi leukocytes {Lou} and residual tissue (R}caeulate
from the morphometric measurements performed at level; rpean vale x)
and standard deviation (8.0)
Tost Control bit TC
x Si. x SD x SD SE
%Co 72 31 «109k 38g
%Y 64 1931-09251
R08) oOT 7-48
%leu 08 «04 «19 OF -13 a
WR «48 «1519 BT ae
‘The difference between the implant (Test! and the tooth (Contrlh; mean
valu (x, standard deviation SD.) and standard error (SE. Difloronc with
96% confidence int ()
86
‘Table 3. % volume ofthe connective tissue components in the area located
adjacent to the interface between the connective tissuefitanium implant
‘and root cementum; collagen (Co), vascular structures (V), fibroblasts (Fi,
leukocytes (Leu) and residue tissue (R cleulatd from the morphometric
‘measurements performed at evel 4; mean value [xl and standard deviation
6D
Test
x $0 x SD » 8D SE
Co 0-23 HOH
HV 83 14-2857 17st
SR 3120S B 2d 13
%lu 03 02-0 «02-0102
HR 4415182051081
‘The diference between the implant (Test) and the tooth (Control); mean
value (2, standard deviation (SD. and standard errr (SE). Difference
‘within 95% confence limit
The soft tissue adjacent to teeth and implant
sites, however, also differed in several respects.
Thus, the connective tissue in specimens from the
implant site harbored a dense network of collagen
fibers which, in major bundles. extended from the
alveolar bone crest to the gingival margin. Most of
the major fiber bundles were aligned in a direction
parallel to the implant surface (Fig. 9)
In the specimens obtained from the tooth sites.
the collagen fiber bundles of the supracrestal con-
nective tissue could be distinguished into (i) dento-
gingival fibers, extending from the supraalveolar
root cementum to the marginal gingival connective
tissue, (ii) dento-periostal fibers, extending from
the root cementum to the alveolar bone crest.
(iii) circular fibers present in the connective tissue
of the marginal gingiva and the supra alveolar con-
nective tissue (Fig. 8)
1m Titanium
i Tooth
co v Fi Leu R
Fig. 10, Histograms describing the % volume of the connective
tissue components in the area located immediately below the
junetional epithelium (area 1); collagen (Co). vascular struc-
tures (V), fibroblasts (Fi), leukocytes (Leu) and residual tissue
(R) of implant and tooth calculated from the morphometric
‘measurements performed at level 4on
in
5
D ete
B
3 a wi
Fig. 11. Histograms illustrating the % volume of the connec:
tive tissue components in the area located adjacent to the
interface between the connective tissuertitanium implant and
root cementum: collagen (Co). vascular structures (V). fibro
blasts (Fi) leukocytes (Leu) and residual tissue (R) of implant
aand tooth caleulated from the morphometric measurements
performed at level 4
Histometric measurements
The results of the histometric measurements are
reported in Table 1. The overall height of the su-
pra crestal soft tissue (GM-BC) was at the implant
site 3.80 mm and at the tooth site (the gingiva)
3.17 mm. While the variation of the height of the
buccal gingiva was small (S.D. 0.04 mm), the soft
tissue adjacent to titanium implants varied be-
tween 3.42 mm and 4.76 mm (S.D. 0.65 mm). The
average height of the junctional epithelium (GM-
aJE) was at the tooth site 2.05 mm compared to
2.14 mm at the implant site. This difference in the
length of the junctional epithelium was not found
to be statistically significant. The length of the
supraalveolar connective tissue, i.e., aJE-BC, was
1,66 mm at the implant site. The corresponding
distance at the tooth site was 1.12 mm. Thus, while
the portion of the soft tissue interface that carried
a junctional epithelium was of similar height in the
two categories of specimen, a significant difference
between the two units was observed with respect to
the length of the supraalveolar connective tissue.
Morphometric measurements
‘The results from the morphometric measurements
of the soft tissue adjacent to implants and teeth are
reported in Tables 2,3 and Figures 10,11, The
connective tissue located immediately lateral to
the junctional epithelium (Area 1) (Table 2.
Fig. 10) was at the tooth site comprised of about
63% collagen fibers, 16% fibroblasts, 7% vascular
structures, 2% inflammatory cells and 12% resid-
ual tissue. In Area 2 (Table 3, Fig. 11) from the
tooth site, there was about 76% collagen, 5% fi-
Soft tissue at implants and teeth
broblasts, 3% vascular structures. 0.5% inflamma-
tory cells and 15% residual tissue. It is obvious that,
the connective tissue of the free gingiva immedi-
ately lateral to the junctional epithelium contained
a larger volume of fibroblasts, but a smaller vol-
ume of collagen and vascular structures than the
supraalveolar connective tissue lateral to the root
cementum (Area 1 versus Area 2)
‘The corresponding connective tissue areas at the
implant sites had a more uniform composition
Thus, the collagen content, the amount of vascular
structures, inflammatory cells and residual tissue
were close to similar in Areas 1 and 2. In Area 1
from the implant sites, however. the volume of
collagen was larger and the amount of fibroblasts.
inflammatory cells and residual tissue smaller than,
in the corresponding area from the tooth sites. The
differences between the two sites with respect to
collagen and fibroblasts persisted but were less
conspicuous in Area 2.
ussion
The present inve: n, performed in dogs
maintained in a careful plaque control program,
demonstrated that the soft tissue. the periimplant
mucosa, that formed at titanium implants follow-
ing abutment connection had many features in
common with gingival tissue at teeth. Thus, like
the gingiva the periimplant mucosa established a
cuff-like barrier (seal) which clinically (i) seemed
to adhere to the surface of the titanium abutment
and (ii) failed to bleed on gentle probing, Further.
a microscopic examination of the two soft tissue
units disclosed that both varieties had a well ker-
atinized oral epithelium which was continuous with
a junctional epithelium that faced the enamel or
the titanium surface. While a majority of the colla-
gen fiber bundles of the free gingival units exa
ined originated from the root cementum, in the
mucosa at implants they appeared to commence
from the marginal bone crest and were orient
parallel with the abutment surface. All ginj
and periimplant units examined in this model ex-
periment were free from infiltrates of inflamma-
tory cells. There are reasons to suggest, therefore,
that under the conditions of study both kinds of
soft tissue units have a proper potential to prevent,
subgingival plaque formation.
In the present study, the titanium implants were
installed in the alveolar process and the coronal
portion of the fixtures were, following surgery,
covered by the mucosa of the edentulous alveolar
ridge. Thus, the tissue which after abutment con-
nection formed the soft periimplant cuff originated
from masticatory mucosa. The histological exam-
ination of the gingiva and the periimplant mucosa
87Berglundh et al.
demonstrated that there were marked differences
between the two units with respect to the orien-
tation of the collagen fibers. In the gingiva, the
collagen fibers were (i) attached to the root ce-
mentum and advanced in perpendicular direction
into the lateral portions of the soft tissue, or
ii) occurred as circular fibers without attachment
to the tooth. In the periimplant tissue, the vast
majority of the coarse fiber bundles were attached
to the marginal bone, not to the titanium surface,
and mainly arranged in a direction parallel to the
surface of the titanium abutment. This observation
appears to be at variance with data previously re-
ported from similar experiments in the monkey
and dogs (¢.g., Schroeder et al. 1981, Buser et al
1989). In these studies. using titanium implants
(ITI), it was observed that: (i) the collagen fibers
of the periimplant mucosa “inserted into the neck
of the implant and were obviously so firmly
tached...” (Schroeder et al. 1981); (ii) “around the
implants anchored in keratinized mucosa, perpen-
dicularly arranged horizontal connective-tissue fi-
bers firmly attached to the implant surface. were
found prevailing” (Buser et al. 1989). The reason
for this difference in the composition of the peri-
implant mucosa of this study and in the experi-
ments by Schroeder et al. (1981) and Buser et al
(1989) is presently not understood but may be
related to differences inherent in the design of the
implant systems used.
The histometric analysis of the biopsy material
from the present experiment revealed that the pe-
riimplant mucosa as well as the gingival unit har-
bored a junctional epithelium. the length of which
was about 2 mm. The observation that a junctional
epithelium formed at the interface between the
titanium surface and the periimplant mucosa is,
consistent with findings previously reported by
Gould et al. (1981, 1984), Jansen et al. (1985),
McKinney et al. (1988) and Van Drie et al. (1988),
‘The formation of a junctional epithelium adjacent
to implants is according to Ten Cate (1985) a log-
ical result of wound healing. Thus, the epithelial
cells proliferate and migrate on the exposed con-
nective tissue surface until the epithelial continuity
is restored.
At both the tooth and implant sites, the apical
cells of the junctional epithelium (IE) terminated
about I-1.5 mm coronal of the alveolar bone crest
At the dentogingival region, the presence of a root
cementum apical of the cementoenamel junction
(CEJ) explains why under normal conditions the
junctional epithelium terminates at or close to
CEI. The reason why at the implant sites the JE
failed to reach the bone crest is not properly un-
derstood. Obviously, however, during healing, the
interface between the supraalveolar portion of the
88
titanium susface and the mucosa must not be rec~
ognized as “exposed” connective tissue. There are
reasons to suggest. therefore, that an interaction
must occur between the surface layer of the tita-
nium implant and the connective tissue. and that
this interaction or “connective tissue integration”
may prevent further epithelial migration. Such a
ive tissue integration” must occur during
I phase of wound healing after abutment
connection and may establish an important biolog-
ical barrier to ensure implant success.
‘A morphological analysis of the tissue immedi-
ately adjacent to bulk metal implants is hampered
by the difficulty to obtain thin sections (