Clin Oral Impl Res 2000: 11: 448–457 Copyright C Munksgaard 2000

Printed in Denmark ¡ All rights reserved

ISSN 0905-7161

An in vitro study on compensation of mismatch

of screw versus cement-retained implant
supported fixed prostheses
Pietrabissa R, Gionso L, Quaglini V, Di Martino E, Simion M. An in Riccardo Pietrabissa1,
vitro study on compensation of mismatch of screw versus cement-retain- Luca Gionso2,
ed implant supported fixed prostheses. Virginio Quaglini3,
Clin Oral Impl Res 2000: 11: 448–457. C Munksgaard 2000. Elena Di Martino4,
Massimo Simion2
In common practice a perfect fit of the prosthetic framework with the 1
Laboratory of Biological Structure
implant abutments is almost impossible to achieve. The mismatch, which Mechanics, Dipartimento di
is principally induced by the technological process adopted to manufac- Bioingegneria, Politecnico di Milano,
ture the fixed prostheses, strains the framework thus generating constraint Milano; 2Università degli Studi di Milano,
reactions. These are static forces that load the implant components and Servizio di Odontostomatologia I,
the bone at the implant–bone interface and may cause the bone remodel- Ospedale San Raffaele, Milano;
3
Laboratory of Biological Structure
ling. Depending on the magnitude of such forces, i.e. depending on the
Mechanics, Dipartimento di Ingegneria
magnitude of the mismatch, the bone remodelling may lead to the Strutturale, Politecnico di Milano, Milano;
loosening of the screws and of the implant–bone interface and hence cause 4
CeBITeC, Dipartimento di
implant failure. The present study shows an in vitro comparison of 3 differ- Bioingegneria, Politecnico di Milano and
ent connecting abutments (standard, EsthetiCone and CerAdapt, Nobel IRCCS San Raffaele, Milano, Italy
Biocare AB, Göteborg, Sweden) with relation to the mechanical stresses
induced by geometrical mismatches (technology induced errors). Two ex-
perimental devices were purposely realized and used to assess the ability
of the different abutments to compensate errors. One was designed for Key words: dental implants – fixed
translation errors and the other for rotation errors. The experimental prosthesis mismatching – position
apparatus set-up includes 2 freestanding implants supporting a prosthetic errors – experimental study –
structure and the connecting abutments. The implants and the abutments biomechanics
were used as delivered by the manufacturer, while the prostheses were
purposely realized and instrumented with strain gauges. The data obtained Prof. Riccardo Pietrabissa, Laboratory of
Biological Structure Mechanics,
with the error devices do not give quantitative information on what hap-
Dipartimento di Bioingegneria,
pens in clinical applications where the implants are connected to living Politecnico di Milano, Piazza L. da Vinci,
bone, which is a tissue much more deformable than the steel used for the 32, 20133 Milano, Italy
error devices. Results allow direct comparisons of the behaviour of the Tel.: π39 02 2399 3349
different investigated abutments with respect to position errors. The Cer- Fax: π39 02 2399 3360
Adapt system (cement retained ceramic abutments) showed the least e-mail pietrabissa/biomed.polimi.it
strain in presence of translation errors. The standard system (screw retain-
ed abutments) showed the least strain in presence of rotation errors. Accepted for publication 23 August 1999

The use of osseointegrated implants as a support
for dental prostheses has obtained general consent
by most clinicians worldwide. The Brånemark im-
plant technique was originally introduced for
treatment of completely edentulous jaws and the
prosthetic components consequently have been de-
signed in order to satisfy this purpose. Successful
long-term prognosis of osseointegrated Brånemark
oral implants in totally edentulous jaws (Adell et
al. 1981; Albrektsson 1988; Albrektsson & Senner-
by 1991) has encouraged several clinicians to ex-
tend the application of dental implant treatment
to partial edentulism. In the last few years, several
studies have shown promising results in the treat-
ment of partially edentulous patients with both
fixed bridge restorations and single tooth replace-

448
 Compensation of mismatch of implant supported fixed prostheses
ments (Jemt 1986; Jemt et al. 1989; Jemt et al.
1992; van Steenberghe 1989; van Steenberghe et al.
1991; Henry et al. 1993).
Recently, the application of dental implants to
partial fixed prosthetic restorations has increased
the esthetic requirements of the final restoration.
Therefore, new abutment designs have been intro-
duced in clinical practice as alternatives to the
standard abutment cylinder. A conical abutment
(EsthetiCone, Nobel Biocare AB, Göteborg,
Sweden) is available since 1991 for multiple unit
reconstruction. The advantage of this abutment
design is that it is available in a 1 mm height con-
figuration, which allows the prosthetic finishing
line to be positioned under the gingival margins.
This characteristic prevents non-esthetic titanium
abutment exposure over the gingiva even in cases
in which thin mucosal tissues are present. In 1989
the traditional concept of prosthetic superstruc-
tures connected to implants through transmucosal
abutments and retained with screws was modified
with a new anchoring method. This method, using
the so called ‘‘UCLA’’ abutment (Lewis et al.
1988) provides the fabrication of customized gold
alloy abutments directly connected to the implant
heads. The prosthetic superstructure is temporarily
cemented on the gold abutments with the same
technique used for cementation on copings in
natural tooth abutments. Cement-retained pros-
theses benefit from the absence of the non-esthetic
screw access through the crown surface, with the
possibility to adapt the shape of the crown and the
emerging profiles to the specific anatomic situation
and from the positioning of the finishing lines ac-
cording to the shape and location of the mucosal
margins.
Recently, Prestipino & Ingber (1993) have pro-
posed a new ceramic abutment (CerAdapt, Nobel
Biocare AB, Göteborg, Sweden). This particular
abutment consists of a sinterized alumina cylinder
that has to be directly connected to the implant
head prepared like a natural tooth. The prosthetic
superstructure is retained with temporary cement.
The advantages of this abutment are the high bioc-
ompatibility of the material, the improved esthetics
due to the brightness through the mucosal tissues,
the finishing lines and the emerging profile cus-
tomization, and the elimination of the screw holes
through the crown surface. However, the drawback
of all the cement-retained prostheses is the trouble-
some retrievability of the superstructure.
Another issue to be considered in fixed implant
supported prostheses is the possible mechanical
stresses introduced in the system by small dimen-
sional errors in the construction of the superstruc-
ture. The final prosthetic superstructure is manu-
factured through different steps: impression of the
abutment construction, preparation of the master
model, waxing, casting, welding and ceramic pro-
cessing. Each of these steps may add errors result-
ing in a mismatch between the position of the abut-
ments with respect to the fixed prosthesis gold cyl-
inders. Such errors can be related to the horizontal
and vertical position and to the rotation between
the different structures. These errors are defined as
position errors.
Clinicians experience the mismatch due to posi-
tion errors during the fixed prosthesis application
as the mismatch does not allow a smooth assembly
of the fixed prosthesis on the abutments. The mis-
match strains the assembled framework thus gen-
erating reactive forces at the implant–bone inter-
face. These static forces load the bone, and bone
remodelling may occur resulting in implant site
displacement toward the direction which produces
the decrease of the framework strain. The greater
the mismatch, the greater the force applied to the
implant–bone interface, the more pronounced the
bone remodelling process, the higher the risk of
loosening of the implant–bone interface mechan-
ical stability and hence the risk of implant failure.
Lack of data exists about the ability of cement
or screw-retained prosthesis to transfer or compen-
sate the stresses induced by position errors.
This paper focuses on an in vitro study com-
paring 3 different abutment systems with respect to
their compliance when increasing mismatches are
applied.
Materials and methods
Assumptions
The work takes into consideration fixed prosthetic
units bonded and supported by two dental im-
plants. Position errors that cause the mismatch be-
tween the two fixed prosthesis gold cylinders and
the two abutments are six in an absolute Cartesian
coordinate system X, Y, Z: translation errors AX,
AY, AZ and rotation errors qX, qY, qZ. In our case
it is convenient to adopt the fixed prosthesis Car-
tesian coordinate system in which the Z axis is the
axis of one of the two implants and the X axis is
the fixed prosthesis longitudinal axis so that the
XZ plane contains the axis of the two implants if
they are parallel each other. Using this coordinate
system no translation position error may be re-
ferred to the Y axis as shown in Fig. 1a. The Y
error AY means the fixed prosthesis reorientation
and hence the coordinate system reorientation (Xø,
Yø, Z) with only an Xø error (AXø).
Looking at the rotation errors considered as ro-
tations of the Cartesian coordinate system of one
fixed prosthesis gold cylinder with respect to the
other (Fig. 1b), the qZ error may be neglected as
449
Pietrabissa et al.

this paper: two translation errors – AX, called fixed

prosthesis length error, and AZ, called fixed pros-
thesis shear error – and two rotation errors – qX,
called fixed prosthesis torsion error, and qY, called
fixed prosthesis bending error.
All the errors are fixed prosthesis errors under
the hypothesis that they depend on the technologi-
cal process but, as the errors are mismatches be-
tween the fixed prosthesis gold cylinders and the
abutments, hence relative errors, they have been
represented as abutment errors (ĀX, ĀZ, q̄X, q̄Y)
for the experimental reasons explained in the fol-
lowing. Each abutment position error is in the op-
posite direction of the corresponding fixed pros-
thesis position error. It holds: ĀXΩªAX, ĀZΩ
ªAZ, q̄XΩªqX, q̄YΩªqY. The four abutment
errors considered in this work are represented in
Fig. 2.

Fig. 1. (a) The adopted Cartesian coordinate system used to

identify the translation errors. The X axis is the fixed prosthesis
longitudinal axis and the Z axis is the abutment axis. The errors
in Y direction are converted in Xø errors as a consequence of
the fixed prosthesis and coordinate system rotation (angle a)
around the Z axis. (b) Rotation errors are defined as relative
rotations of the two fixed prosthesis gold cylinder coordinate
systems (X1, Y1, Z1 and X2, Y2, Z2). In this occurrence rotation
around X2 or Y2 axis causes qX or qY errors while rotation
around the Z2 axis does not cause errors as the connection site
is supposed to have an axial symmetry.

Fig. 2. Schematic representation of a mandible with two im-

plants supporting a fixed prosthesis. In detail the possible abut-
ment position errors are (a) length error, (b) shear error, (c)
torsion error and (d) bending error.

the rotation of a gold cylinder around its axis of

symmetry does not cause mismatches.
As a consequence of the previously mentioned Fig. 3. (a) Sketch of the device used to impose translation
considerations only four errors are considered in errors. (b) Sketch of the device used to impose rotation errors.

450
 Compensation of mismatch of implant supported fixed prostheses

Experimental procedure
The adopted experimental procedure is intended
to apply the desired abutment position errors and
measure the correspondent fixed prosthesis strains.
Two purposely designed devices have been fabri-
cated and used, a translation error device (Fig. 3a)
and a rotation error device (Fig. 3b). Both the de-
vices have been built in steel.
The translation error device has two slides that
may be translated using two micrometer screws,
0.5 mm pitch. One enables to set ĀX errors and
the other ĀZ errors. The rotation error device has
two shafts that may be rotated to the desired angle
using a goniometer. One shaft enables to set q̄X
errors and the other q̄Y errors. Each slide/shaft of
both devices houses one implant with its abutment.
The fixed prosthesis is equipped with two strain
gauges on its upper surface in order to measure
the strains induced by position errors. The fixed
prosthesis is mounted on the abutments of the
translation (rotation) error device and the slides
(shafts) are translated (rotated) to obtain the mini-
mum fixed prosthesis strain which corresponds to
the minimum position error. In general a global
strain equal to zero cannot be obtained because Fig. 4. The fixed prosthesis instrumented for measuring strains
the translation (rotation) error device cannot com- due to (a) length errors, (b) shear errors, (c) bending errors and
(d) torsion errors.
pensate rotation (translation) errors. Hence the
minimum fixed prosthesis strain is assumed as ref-
erence for abutment position errors equal to zero. for the different error type. As a consequence the
When the minimum position error is reached the strain gauges have been applied in different ways
fixed prosthesis is removed from the abutments and depending on the error type to measure as follows.
the abutments are moved one at a time to set the For fixed prosthesis length and shear errors the
desired error. The fixed prosthesis is mounted strain gauges are both parallel to the fixed pros-
again on the abutments and the strain gauges allow thesis longitudinal axis (X) and they measure
measuring the strains induced by the set position strains in X direction (eX). With reference to Fig.
error. The fixed prosthesis is then removed again 4a and 4b, where the positive direction of the abut-
and the procedure is repeated using translation ment errors are indicated, one notes that:
steps of 20 mm and rotation steps of 1æ. The maxi-
eX1±0 and eX2±0 for ĀX±0
mum investigated translation errors have been set
eX1∞0 and eX2±0 for ĀZ±0.
to 0.1 mm while the maximum investigated ro-
tation errors depend on the feasibility of assemb- For fixed prosthesis bending and torsion errors
ling the fixed prosthesis on the rotated abutments. one strain-gauge is parallel to the fixed prosthesis
longitudinal axis (X) and it measures strains in X
direction (eX) while the other is set to 45æ with re-
Measurements spect to X axis and it measures strains in that di-
Position errors are set moving the translation slides rection (e45). With reference to Fig. 4c and 4d,
or the rotation shafts and measuring the new posi- where the positive directions of the abutment
tion of the moved abutment using a thread gauge errors are indicated, one notes that:
or a goniometer.
eX±0 and e45∑0 for q̄Y±0
The fixed prosthesis strain induced by the posi-
eX∑0 and e45±0 for q̄X±0.
tion error is measured using two strain gauges ap-
plied on the upper surface of the fixed prosthesis. The strain-gauges fixed as previously shown en-
On the basis of the theory of elasticity and con- able to compare the effects due to the same posi-
sidering the fixed prosthesis, as a first approxi- tion errors on fixed prostheses mounted using dif-
mation, similar to an elastic beam, it is possible to ferent connecting abutments as well as the effects
predict the deformed shape of the fixed prosthesis due to increasing position errors.

451
Pietrabissa et al.
Fig. 5. Typical strains measured by each strain gauge as functions of the different position errors.
The strain-gauges used for measurements are
LY13–0.6/120 (Hottinger Baldwin Messtechnik
GmbH, Darmstadt, Germany) with 0.6 mm refer-
ence length in the measurement direction and 120
ohm reference resistance. The strain gauges have
been connected to a conditioner-amplifier
DMCPLUS12 (Hottinger Baldwin Messtechnik
GmbH, Darmstadt, Germany).
Fig. 5 shows typical strains measured by the
strain gauges as functions of the position errors.
Fixed prostheses and implant systems
Six fixed prostheses have been prepared identical
in shape and dimensions. The material used for the
452
fixed prosthesis fabrication has been the V Classic
gold alloy (Metaux Precieux SA Metalor, Neuch-
atel, Switzerland). The fixed prostheses have 20
mm distance between the two gold cylinder axes.
Two Brånemark standard fixtures SDCA 001
(3.75¿10 mm) have been screwed into the slides/
shafts of the error devices to support the abut-
ments.
Three different pairs of Brånemark abutments
have been used to connect the fixed prosthesis to
the implants. The first pair has been the standard
abutment 3 mm high and the second the Estheti-
Cone abutment 3 mm high. In both cases the fixed
prosthesis has been connected using gold screws
according to the indications of the manufacturer.
 Compensation of mismatch of implant supported fixed prostheses
Table 1. Types and dimensions of the tested abutments

Abutment type Junction Height Implant diameter

(mm) (mm)

Standard gold screws 3 3.75

EsthetiCone gold screws 4.75 3.75
CerAdapt cement 11.9 3.75

The screws have been tightened with a Torque

Controller handpiece (Nobel Biocare AB, Göte-
borg, Sweden) using a torque of 20 N cm. The
third type has been the CerAdapt abutment pre-
pared with a high-speed diamond burr. The fixed
prosthesis shoulder has been located 3 mm over
the abutment base and the axial walls have been
gently grided to achieve 6æ of inclination. Before
waxing, a die-spacer layer 20 mm thick has been
applied to each abutment.
The fixed prosthesis has been retained with a
temporary cement (Temp Bond, Kerr Ltd,
Peterborough, United Kingdom) applying an axial Fig. 6. Behaviour of the 3 tested abutments in terms of fixed
force of 100 N for 60 s. The characteristics of the prosthesis strains measured with the translation error device
3 types of abutment are shown in Table 1. imposing abutment length errors.

Results
Figs 6 to 9 show the strains measured on the fixed
prosthesis versus the position errors for the 3
tested connecting abutments. The test procedure
has been evaluated, in a series of repetitions, to
give results with a measuring error lower than 5%
of the measured value.

Abutment length error

In Fig. 6 the fixed prosthesis strains have been cal-
eX1πeX2
culated as . The minimum strains have
2
been obtained using the CerAdapt abutments (the
strain is always under 100 me). In comparison with
the standard abutments, the EsthetiCone abut-
ments induced lower strains for ĀXÆ60 mm and
higher strains for ĀXØ80 mm. The maximum
strain has been 200 me reached using the Estheti-
Cone abutments for ĀXΩ100 mm.

Abutment shear error
In Fig. 7 the strains have been calculated as
eX2ªeX1
. The minimum strains have been ob-
2
tained using the CerAdapt abutments (the strain is
always under 10 me except for ĀZΩ40 mm which
scored a strain of 18 me). In comparison with the
EsthetiCone abutments, the standard abutments
Fig. 7. Behaviour of the 3 tested abutments in terms of fixed
prosthesis strains measured with the translation error device
imposing abutment shear errors.
induced lower strains for ĀZÆ60 mm and higher
strains for greater errors (Ø80 mm). The maximum
absolute strain has been 68 me reached using the
standard abutments for ĀZΩ100 mm.

453
Pietrabissa et al.
Fig. 8. Behaviour of the 3 tested abutments in terms of fixed
prosthesis strains measured with the rotation error device-im-
posing abutment bending errors.
Fig. 9. Behaviour of the 3 tested abutments in terms of fixed
prosthesis strains measured with the rotation error device im-
posing abutment torsion errors.
Abutment bending error
In Fig. 8 the fixed prosthesis strain is eX. The stan-
dard abutments induced the minimum strain for
every imposed bending error and the strain value
(which was always lower than 40 me) has been half
454
or even lower than using the EsthetiCone and Cer-
Adapt abutments. The EsthetiCone abutments in-
duced lower strains than the CerAdapt abutments
except for 3æ bending error. The maximum strain
has been 160 me reached using the EsthetiCone
abutments for q̄YΩ3æ.
Abutment torsion errors
In Fig. 9 the fixed prosthesis strain is e45. The stan-
dard abutments induced the minimum strain for
every torsion error investigated, except for 0 and 1æ.
The strains obtained using these abutments have al-
ways been lower than 70 me. The EsthetiCone abut-
ments induced always the greatest strain with a
maximum of 170 me for q̄XΩ3æ. The CerAdapt abut-
ments induced intermediate values of strain.
It should be noted that the zero strain has not
been achieved with the angular device, i.e. a non-
zero strain is measured by the strain gauges when
no error is set on the device. This is due to the
presence of either fixed prosthesis length or shear
error that cannot be compensated by the rotation
error device.
Discussion
This experimental study was designed in order to
compare 3 different implant–prosthesis connection
systems (one cemented and two screw-retained)
with respect to their capability to compensate fixed
prosthesis dimensional and/or shape errors.
The results from the study showed that the ce-
mented abutments (CerAdapt) have the best capa-
bility of compensating translation (axial and trans-
versal) errors. In fact, at the maximum translation
error (0.1 mm) CerAdapt connected prosthesis in-
duced only half of the strain as compared to both
the prostheses connected with standard or Estheti-
Cone abutments (90 versus 200 me). A possible ex-
planation of this finding could be that the presence
of a deformable cement layer might compensate
the translation errors produced during the manu-
facturing process. The cement could act as an ad-
sorber for the deformation caused by the mis-
match and therefore it could be able to preserve
the structure (implant–abutment–prosthesis) from
greater strains. The gap between the abutments
and the prosthesis introduced by the die-spacer,
during waxing and casting, could allow the forma-
tion of a cement layer with different thickness at
the abutment–prosthesis interface. This might
compensate the translation error and reduce the
strain.
The standard and EsthetiCone abutments
showed approximately the same capability of com-
 Compensation of mismatch of implant supported fixed prostheses
pensating translation errors and the strains of the
prosthesis displayed an almost linear increasing
with error. However, shear errors were more easily
compensated (strain lower than 70 me for 0.1 mm
error) in comparison with axial errors (200 me for
0.1 mm error).
With rotation errors, the EsthetiCone abutments
induced lower capability of compensating both
bending and torsion errors. In fact, they caused
greater strains in most of the tests as compared to
both the other two systems. The lower compensat-
ing ability of the EsthetiCone abutments in pres-
ence of rotation errors could be due to their larger
height as compared to the standard abutments.
Major height could cause a greater horizontal dis-
placement of the top of the abutment at the same
degree of rotation.
For rotation errors the behaviour of the CerAd-
apt abutments stands between those of the stan-
dard and EsthetiCone abutments. This behaviour
could be correlated to the combination of two fac-
tors: 1) the overall height of the tested CerAdapt
abutments was the highest and this causes a
stronger lever arm and 2) the compensating effect
of the cement layer. However, the results concern-
ing the CerAdapt abutments should be regarded
with caution as their error transfer capability could
vary depending on the shape and the height of the
final abutment preparation in the different clinical
situations.
The results from this experimental study are
suitable for performance of direct comparisons of
the behaviour of the 3 considered abutments when
connected to prosthetic superstructures having po-
sition errors. The data obtained with the error de-
vices do not give quantitative information on what
happens in clinical applications where the implants
are connected to living bone which is a tissue much
more deformable than the steel used for the error
devices.
The results show that system CerAdapt abut-
ments/cemented prosthesis transfers the least
strain in presence of translation errors, the system
standard abutments/screw retained prosthesis
showed the highest capability of compensating ro-
tation errors, whereas the system EsthetiCone
abutments showed the least capability.
Further studies are necessary to quantify the in-
fluence of position errors on long-term success of
the implant-supported restorations in clinical prac-
tice.
Résumé
En pratique privée, une adaptation parfaite de la prothèse sur
l’implant est pratiquement impossible. Cette difficulté qui est
principalement provoquée par le processus technologique utilisé
pour fabriquer les prothèses fixes, met à l’épreuve la superstruc-
ture et provoque donc des contraintes. Ces forces statiques qui
chargent les composants de l’implant et de l’os au niveau de
l’interface os-implant peuvent donc occasionner le remodelage
osseux. Suivant l’amplitude de ces forces, c.-à-d. suivant la ma-
gnitude de cette disparité, le remodelage osseux peut provoquer
le relâchement des vis et celui de l’interface os-implant provo-
quant alors l’échec. L’étude compare in vitro trois connections
différentes (standard, EsthetiCone et CerAdapt, Nobel Biocare
AB, Göteborg, Suède) en relation avec les stress mécaniques
induits par des disparités géométriques (erreurs induites
technologiquement). Deux systèmes expérimentaux ont été réa-
lisés et utilisés pour vérifier l’habilité des différents piliers à
compenser les erreurs. Un système a été conçu pour les erreurs
de translation et l’autre pour les erreurs de rotation. L’appareil
expérimental comprenait deux implants libres portant une
structure prothétique ainsi que les piliers de connection. Les
implants et les piliers ont été utilisés suivant les indications don-
nées par le fabriquant tandis que les prothèses ont été réalisées
et accompagnées de jauges de force. Les données obtenues avec
les systèmes d’erreur ne donnent pas d’information quantitative
sur ce qui arrive dans les applications cliniques où les implants
sont reliés à l’os vivant qui est un tissu beaucoup plus déforma-
ble que l’acier utilisé pour les systèmes d’erreur. Les résultats
permettent des comparaisons directes dans le comportement
des piliers étudiés vis-à-vis des erreurs de position. Le système
CerAdapt accusait la contrainte la plus faible en présence des
erreurs de translation. Le système standard, c.-à-d. celui dont
les piliers étaient retenus par des vis, possédait la force la plus
faible en présence des erreurs de rotation.
Zusammenfassung
In der Privatpraxis ist es fast unmöglich, eine perfekte Passge-
nauigkeit zwischen dem prothetischen Gerüst und den Implan-
tataufbauteilen zu erreichen. Diese Passungenauigkeit, welche
im Prinzip durch die individuelle Anfertigung der festsitzenden
Rekonstruktion bedingt ist, setzt das Gerüst unter Spannung
und bewirkt zwangsweise Reaktionen. Es handelt sich um stati-
sche Kräfte, welche die Implantatkomponenten und den Kno-
chen an der Implantat-Knochen-Kontaktfläche belasten und zu
Umbauvorgängen im Knochen führen können. Abhängig von
der grösse dieser Kräfte, d.h. abhängig vom Ausmass der Pas-
sungenauigkeit, können die Umbauvorgänge im Knochen zur
Lockerung von Schrauben und zur Lockerung im Bereich der
Knochen-Implantat-Berührungsfläche führen und dadurch ei-
nen Implantatmisserfolg bewirken. Die vorliegende Studie zeigt
einen in vitro Vergleich von 3 verschiedenen Prothetikaufbauten
(Standard, EsthetiCone und CerAdapt, Nobel Biocare AB, Gö-
teborg, Sweden) in Relation zu mechanischem Stress, welcher
durch geometrische Passungenauigkeiten induziert wird (tech-
nisch bedingte Fehler). Es wurden zwei experimentelle Vorrich-
tungen entwickelt und angewendet, um zu testen, inwiefern die
verschiedenen Prothetikteile Fehler kompensieren können. Eine
Vorrichtung wurde für die Messung von Translationsfehlern
und die andere für die Registrierung von Rotationsfehlern ent-
wickelt. Der experimentelle Aufbau bestand aus zwei freistehen-
den Implantaten, welche eine prothetische Rekonstruktion und
prothetische Aufbauteile trugen. Die Implantate und die Auf-
bauteile wurden so verwendet, wie sie vom Hersteller geliefert
wurden, während die Rekonstruktionen individuell angefertigt
und mit Dehnmessstreifen versehen wurden. Die Daten, wel-
sche mit den Fehlervorrichtungen aufgezeichnet wurden, erge-
ben keine quantitative Information über das, was bei der klini-
schen Anwendungen bei Implantaten, die mit vitalem Knochen
verbunden sind, passiert. Bei Knochen handelt es sich um vita-
les Gewebe, welches leichter deformierbar ist als der Stahl, der
für die Fehlervorrichtungen verarbeitet wurde. Die Resutate er-
lauben den direkten Vergleich des Verhaltens von den verschie-
denen untersuchten Prothetikteilen in Bezug auf Positionie-
455
Pietrabissa et al.
rungsfehler. Das CerAdapt System (zementierte keramische
Prothetikteile) zeigte die kleinste Spannung in Bezug auf Trans-
lationsfehler. Das Standardsystem (verschraubte Prothetikteile)
zeigte die kleinste Spannung in Bezug auf Rotationsfehler.

Resumen
En la practica, el ajuste perfecto de la estructura protésica con
los pilares de los implantes es casi imposible de conseguir. El
desajuste, que es inducido principalmente por los procesos tec-
nológicos utilizados para fabricar la prótesis fija, contrae la
estructura generando por ello reacciones a la contracción. Estas
son fuerzas estáticas que cargan los componentes del implante
y el hueso en la interfase hueso-implante y pueden causar remo-
delación ósea. Dependiendo de la magnitud de dicha fuerza,
i.e. dependiendo de la magnitud del desajuste, la remodelación
ósea puede conducir al aflojamiento de los tornillos y de la
interfase hueso-implante y por tanto causar fracaso del implan-
te. El presente estudio muestra la comparación in vitro de 3
pilares de conexión (standard, EsthetiCone y CerAdapt, Nobel
Biocare AB, Goteborg, Sweden) en relación con los estrés mecá-
nicos inducidos por desajustes geométricos (errores inducidos
tecnológicamente). Se realizaron dos dispositivos experimenta-
les a propósito y utilizados para valorar la habilidad de los
diferentes pilares para compensar los errores. Se diseñó uno
para compensar los errores de traslación y otro para los errores
de rotación. El montaje del aparato experimental incluye dos
implantes libres soportando una estructura protésica y sus pila-
res de conexión. Los implantes y sus pilares se usaron tal y
como se distribuyeron por el fabricante, mientras que las próte-
sis fueron fabricadas a propósito e instrumentadas con medido-
res de tensión. Los datos obtenidos con los dispositivos de error
no dan una información cuantitativa de los que ocurre en las
aplicaciones clı́nicas cuando los implantes se conectan a hueso
vivo que es un tejido mucho más deformable que el acero usado
para los dispositivos de error. Los resultados permiten compa-
raciones directas del comportamiento de los diferentes implan-
tes investigados respecto a los errores de posición. El sistema
CerAdapt (pilares cerámicos retenidos por cemento) mostraron
la menor tensión en presencia de errores de traslación. El siste-
References
ma estándar (pilares retenidos por tornillos) mostró la menor Adell, R., Lekholm, U., Rockler, B. & Brånemark, P.I. (1981)
tensión en presencia de errores de rotación. A 15-year study of osseointegrated implants in the treatment
of the edentulous jaw. International Journal of Oral Surgery
6: 387–416.
Albrektsson, T. (1988) A multicenter report on osseointegrated
oral implants. Journal of Prosthetic Dentistry 60: 75–84.
Albrektsson, T. & Sennerby, L. (1991) State of the art in oral
implants. Journal of Clinical Periodontology 18: 474–481.
Henry, P., Tolman, D. & Bolender, C. (1993) The applicability
of osseointegrated implants in the treatment of partially
edentulous patients: three-year results of a prospective multi-
center study. Quintessence International 24: 123–129.
Jemt, T. (1986) Modified single and short-span restorations
supported by osseointegrated fixtures in the partially eden-
tulous jaw. Journal of Prosthetic Dentistry 55: 243–247.
Jemt, T., Lekholm, U. & Adell, R. (1989) Osseointegrated im-
plants in the treatment of partially edentulous patients: a
preliminary study on 876 consecutively placed fixtures. In-
ternational Journal of Oral and Maxillofacial Implants 4:
211–217.
Jemt, T., Lindèn, B. & Lekholm, U. (1992) Failures and compli-
cations in 127 consecutively placed fixed partial prostheses
supported by Brånemark implants: from prosthetic treat-
ment to first annual check-up. International Journal of Oral
and Maxillofacial Implants 7: 40–44.
Lewis, S., Beumer, J.3d, Hornburg, W. & Moy, P. (1988) The

456
 Compensation of mismatch of implant supported fixed prostheses
‘‘UCLA’’ abutment. International Journal of Oral and Max- porting fixed partial prostheses in the treatment of partial
illofacial Implants 3: 183–189. edentulism. Journal of Prosthetic Dentistry 61: 217–223.
Prestipino, V. & Ingber, A. (1993) Esthetic high-strength im- van Steenberghe, D., Brånemark, P.I., Quirynen, M., De Mars,
plant abutments. Part I. Journal of Esthetic Dentistry 5: 29– G. & Naert, I. (1991) The rehabilitation of oral defects by
36. osseointegrated implants. Journal of Clinical Periodontology
van Steenberghe, D. (1989) A retrospective multicenter evalu- 18: 488–493.
ation of the survival rate of osseointegrated fixtures sup-

457