Interface shear strength of titanium implants with a

sandblasted and acid-etched surface: A biomechanical

study in the maxilla of miniature pigs

Daniel Buser,1 Thomas Nydegger,2 Thomas Oxland,2 David L. Cochran,3 Robert K. Schenk,1
Hans Peter Hirt,1 Daniel Snétivy,4 Lutz-Peter Nolte2
1
Department of Oral Surgery, School of Dental Medicine, University of Berne, Freiburgstrasse 7, CH-3010
Berne, Switzerland
2
Maurice E. Müller Institute for Biomechanics, University of Berne, CH-3010 Berne, Switzerland
3
Department of Periodontology, University of Texas Health Science Center, San Antonio, Texas 78284, USA
4
Department of Research and Development, Institut Straumann AG, CH-4437 Waldenburg, Switzerland

Received 18 December 1997; accepted 2 July 1998

Abstract: The purpose of the present study was to evaluate (1.39 vs. 1.14 Nm) without reaching statistical significance.
the interface shear strength of unloaded titanium implants At 8 and 12 weeks of healing, the two rough surfaces
with a sandblasted and acid-etched (SLA) surface in the showed similar mean RTVs. The implant position also had a
maxilla of miniature pigs. The two best documented sur- significant influence on removal torques for each implant
faces in implant dentistry, the machined and the titanium type primarily owing to differences in density in the peri-
plasma-sprayed (TPS) surfaces served as controls. After 4, 8, implant bone structure. It can be concluded that the interface
and 12 weeks of healing, removal torque testing was per- shear strength of titanium implants is significantly influ-
formed to evaluate the interface shear strength of each im- enced by their surface characteristics, since the machined
plant type. The results revealed statistically significant dif- titanium surface demonstrated significantly lower RTV in
ferences between the machined and the two rough titanium the maxilla of miniature pigs compared with the TPS and
surfaces (p < .00001). The machined surface demonstrated SLA surfaces. © 1999 John Wiley & Sons, Inc. J Biomed
mean removal torque values (RTV) between 0.13 and 0.26 Mater Res, 45, 75–83, 1999.
Nm, whereas the RTV of the two rough surfaces ranged
between 1.14 and 1.56 Nm. At 4 weeks of healing, the SLA Key words: interface shear strength; removal torques; SLA
implants yielded a higher mean RTV than the TPS implants surface; titanium implants; TPS surface

INTRODUCTION cessful osseointegration was also reported in clinical

studies using cp titanium implants.6–13 Today, screw-
type titanium implants of various shapes are mainly
In the past 30 years, the use of endosseous dental
preferred in implant dentistry.
implants has become a scientifically accepted and
In recent years, attempts have been made to im-
well-documented treatment modality for fully and
partially edentulous patients. These implants are prove bone anchorage of endosseous dental implants.
placed with precise fit in the jaw bone and require no Thomas and Cook14 examined the variables that po-
cement for stabilization. Fundamental studies con- tentially influence bone apposition to an implant sur-
ducted by the research teams of Brånemark et al.1,2 face. Of 12 parameters studied, only surface character-
and Schroeder et al.3–5 using commercially pure (cp) istics were shown to have a significant effect on bone
titanium implants in animal models showed that im- integration of an implant. This observation was con-
plant anchorage with direct bone contact can be firmed in a histometric study by Buser et al.15 in long
achieved. This direct bone–implant interface is fre- bones of miniature pigs. Of the five tested, unloaded
quently termed osseointegration.2 Subsequently, suc- titanium implants, the sandblasted and acid-etched
surface (SLA) demonstrated the best bone apposition
to the implant surface, with a mean bone–implant con-
Correspondence to: D. Buser tact (BIC) of 52% and 58% at 3 and 6 weeks of healing,
Contract grant sponsor: ITI Foundation for the Promotion
respectively. The next best surface was the titanium
of Oral Implantology; Contract grant number: 94-080
plasma-sprayed (TPS) surface (38% BIC), a clinically
© 1999 John Wiley & Sons, Inc. CCC 0021-9304/99/020075-09 well-documented titanium surface in implant den-
76 BUSER ET AL.

tistry.16 Polished and fine-textured titanium surfaces maxilla of miniature pigs. All implants had a solid-screw
were clearly less effective, with BICs ranging from design and were of identical shape (Institute Straumann AG,
20% to 25%. A parallel biomechanical study by Wilke Waldenburg, Switzerland), with an outer diameter of 4.05
et al.17 tested removal torque values (RTV) of un- mm. The implants had no macroscopic retentive elements
such as a vent in the apical portion or grooves in the long
loaded titanium implants with various surface charac-
axis. The intraosseous portion of the implants was charac-
teristics in the tibia of sheep, and demonstrated a sig-
terized by three different surfaces and measured 8 mm in
nificant difference between rough and polished or length (Fig 1). On top of the implants, a square head was an
fine-textured titanium surfaces. The SLA surface integrated part of the implant to allow direct attachment of
showed the best results, reaching a mean RTV of 6.4 the adapter for implant insertion and measurement of the
Nm at 6 months of healing, followed by the TPS sur- removal torques. The following three different implant types
face with 5.3 Nm. Polished or fine-textured surfaces were tested:
showed a mean RTV of 0.4–0.7 Nm during the course
Type SLA = sandblasted with large grit (0.25–0.50 mm) and
of the study period. These observations were also con-
acid attacked with HCl/H2SO4
firmed in a study by Wong et al.,18 which demon- Type TPS = titanium plasma-sprayed
strated the best results for SLA implants among vari- Type M = machined
ous titanium surfaces without functional load measur-
ing both the percentage of BIC and pull-out forces. The surface characteristics of a typical sample for each of the
These three studies were all carried out in the femur three different surfaces were examined qualitatively by
scanning electron microscopy [Fig. 2(a–c)].
and/or tibia of miniature pigs or sheep; hence, there
To determine the profile quantitatively, profilometry was
was a need to examine this new SLA surface in jaw
performed using a Form Talysurf Series 2 laser interferomet-
bone prior to the clinical testing in patients. In a re- ric system (Rank Taylor Hobson Ltd., Leicester, UK)
cently published study by Cochran et al.,19 the per- equipped with a custom-made 0.6-␮m-diameter diamond
centage of BIC was histometrically examined for SLA stylus. For each implant type, two samples were scanned
implants in the canine mandible. Both unloaded and along the circumference in three or four different areas over
loaded SLA implants were evaluated with healing pe- a length of 2 mm. Thirty-one different amplitude, spacing,
riods up to 15 months, and titanium implants with a and hybrid parameters were calculated from the profile
TPS surface served as controls. The study demon- data. Average roughness Ra and the mean spacing of adja-
strated a significantly better bone apposition for SLA cent local peaks S were selected as the variables to best
implants at 3 and 15 months of healing, whereas no describe the surface characteristics.
Ra is a universally recognized and often used parameter
significant differences were found between the im-
describing roughness. It is the arithmetic mean of the depar-
plant surfaces at 6 months of healing.
tures of the roughness profile from the mean line, and is
The purpose of the present study was to test this calculated as:
promising titanium surface for the first time biome-
chanically in jaw bone evaluating the interface shear
兰 |y共x兲|dx
1 L
Ra =
strength of SLA implants in the maxilla of miniature L 0

pigs. The two best documented titanium surfaces in

where L is the assessment length and y(x) is the profile am-
implant dentistry, the machined and the TPS surface,
plitude. S is the mean spacing of adjacent local peaks, mea-
served as controls. The tested hypothesis was that the
sured over the assessment length, and is calculated as:
SLA surface would reveal higher removal torque val-
ues compared with the control surfaces.

MATERIALS AND METHODS

Approval for animal research

The protocol of the animal study was approved by the

standing committee on Animal Research at the University
Hospital, Medical Faculty, University of Berne, and by the
Committee for Animal Research, State of Berne (Approval
No. 96/95). The guidelines of the State of Berne for the care
and use of laboratory animals were followed.

Implant design and implant surfaces

Figure 1. The three tested titanium screw-type implants
Commercially pure titanium implants with three different with the same macroscopic shape, but different surfaces
surfaces were placed in edentulous areas of the anterior (f.l.t.r.): SLA, TPS, and machined surface.
REMOVAL TORQUE VALUES OF SLA IMPLANTS 77

Surgical procedures

A total of nine adult miniature pigs with a minimum age

of 2 years were used in the study. In each animal, two sur-
gical interventions were performed. During the first surgical
procedure, the anterior teeth in the maxilla were removed
under general anesthesia using extended mucoperiosteal
flaps to provide sufficient access to the alveolar crest con-
taining the teeth to be removed (Surgical Research Unit ESI
and Clinic for Large Animals, University of Berne). Prior to
surgery, the animals were given 1 g of prophylactic amoxi-
cillin intramuscularly. Following tooth removal, the el-
evated flaps were repositioned and sutured with interrupted
sutures. Following a healing period of at least 6 months, a
(a) total of 54 implants were inserted, 6 in each animal. The
recipient sites in the created edentulous areas of the maxilla
were exposed by the elevation of buccal mucoperiosteal
flaps. When necessary, the alveolar crest was flattened to
allow for precise preparation of the implant recipient sites,
using three different precise spiral drills of increasing diam-
eter at 500 rpm and copious irrigation with chilled sterile
physiological saline. Subsequently, the thread was cut into
the bone cavity with a tap. Thus, primary stability was
achieved by the press-fit design and the screw thread. Three
implants were inserted on each side of the maxilla, one of
each type (Fig. 3). The anterior-posterior position of each
implant type was varied in each animal to give no prefer-
ence for one particular implant type. Insertion torques were
measured using a custom-designed, sterilizable torque
wrench instrumented with strain gauges. Primary wound
(b) closure was achieved with resorbable suture material.

Removal torque testing

The miniature pigs were sacrificed after three different

healing periods of 4, 8, or 12 weeks. Each subgroup con-
sisted of three miniature pigs with six implants each. Imme-
diately after sacrifice, the soft tissues in the edentulous areas
of the maxilla were removed to expose the integrated tita-
nium implants. Subsequently, the maxilla was excised and
the left and right halves were isolated with a diamond
plated saw (Makro Trennsystem; Exakt Apparatebau AG,
Norderstedt, Germany). To improve further handling and
for temperature isolation, each of the samples was molded
(c) into dental cement (Kerr Suprastone Green; Kerr Europe
AG, Basel, Switzerland).
Figure 2. (a–c) SEM pictures of the three tested surfaces The removal torque testing was performed on a biaxial
(f.l.t.r.): SLA, TPS, and M surface. hydraulic materials testing machine (MTS Minibionix
358.02; Minneapolis, MN). To apply pure axial moments in
the test, the axis of the dental implants had to correspond to
the axis of the testing machine. For this reason, the implant
S1 + S2 + … + Sn
n
1
S=
n 兺S =
i=1
i
n
was first attached to the actuator, thereby guaranteeing the
implant-actuator alignment. The implant–bone–dental ce-
ment complex was lowered into a tub on the rigid part of the
where Si is the spacing between two local peaks and n is the machine, which was then filled with low-temperature melt-
number of spacings over the assessment length. A local peak ing metal alloy (Legierung 47° Grad; Billiton Witmetaal B.V.,
is the highest part of the profile measured between two ad- Naarden, The Netherlands). The cooling of the alloy effec-
jacent minima. tively fixed the implant–bone–dental cement complex to the
78 BUSER ET AL.

Figure 4. Typical graph of a removal test. The peak of the

curve was deemed the failure torque of the bone–implant
interface.

characterize the bone–implant interface, the removal torque

was defined as the maximum torque on the curve.

Histologic and histomorphometric analysis

Following removal torque testing, the block specimens in-

cluding the implants were placed in 4% formalin/1% CaCl2
fixative. The specimens were dehydrated and embedded in
methylmethacrylate. They were cut buccal-lingual in the
long axis, allowing one to three undecalcified sections ∼500
␮m in thickness per implant, using a low-speed diamond
saw with coolant. Subsequently, the sections were glued
with acrylic cement to opaque Plexiglas, ground to a final
thickness of ∼80 ␮m, and stained superficially with Tolu-
idine blue combined with basic fuchsin.20 In addition to the
histologic evaluation, the central specimen of each implant
was analyzed histometrically by an experienced examiner.
The mean volume of bone matrix of the periimplant bone
structure within 2 mm from the implant surface was mea-
sured using a grid for point counting at a magnification of
×100 to evaluate the bone density in this area.

Figure 3. (a–c) Typical profilometric graphs of the quanti-

tative surface roughness measurements (profilometry) for Statistical analysis
the SLA, TPS, and M surfaces.
The peak insertion torque was analyzed for implant type
machine (Fig. 4). The removal torque test was performed by and position using a two-factor analysis of variance
applying a counterclockwise rotation to the implant axis at a (ANOVA). The removal torque and histological data were
rate of 0.1°/s. analyzed for implant type, healing period, and implant po-
After testing one implant, the potting alloy was melted to sition using a three-factor ANOVA. Post-hoc pairwise com-
remove the implant–bone–dental cement complex from the parisons were made using Scheffé’s F test. Assumptions im-
fitting tub. The next implant was secured to the actuator and plicit in ANOVA such as normally distributed groups and
the entire process repeated until all implants were tested. To homogeneity of variance were tested. All analyses were
avoid drying the bone, the specimens were sprayed with computed with the ANOVA/MANOVA module of Statis-
saline solution every 15 min. tica for Windows, Release 5.1 (StatSoft, Tulsa OK). A type I
For all tests, the torque-rotation curve was recorded. To error of .05 was deemed significant.
REMOVAL TORQUE VALUES OF SLA IMPLANTS 79

RESULTS

The profiles of the three different surfaces are

shown in Figure 3. The microporous TPS surface dem-
onstrated the highest values for average roughness
(average Ra = 3.1 ␮m) and for mean spacing of adja-
cent local peaks (average S = 17.8 ␮m). The SLA type
showed slightly lower values (Ra = 2.0 ␮m; S = 12.0
␮m), whereas the machined surface was clearly found
to have the lowest roughness parameters (Ra = 0.15
␮m; S = 9.6 ␮m).
The peak insertion torques ranged from 0.07 to 0.66
Nm. There was no significant effect of position (p <
.12) and a marginally significant effect of implant type
(p < .06). The machined implants showed the highest Figure 5. Removal torque values of the three implant types
insertion torques (0.31 ± 0.12 Nm), followed by the after 4, 8, and 12 weeks of healing
TPS (0.28 ± 0.10 Nm) and SLA implants (0.20 ± 0.12
Nm). After insertion, all implants exhibited primary
stability. surfaces. The direct comparison between the TPS and
At sacrifice and following soft-tissue removal, two SLA implants demonstrated a higher mean RTV for
implants exhibited mobility and one additional type SLA implants at 4 weeks of healing (1.39 vs. 1.14 Nm)
M implant demonstrated a deep bony crater. These without reaching statistical significance. At 8 and 12
three implants were consequently excluded from fur- weeks of healing, both surfaces showed similar re-
ther analysis. The remaining 51 implants demon- sults.
strated a firm anchorage in the maxillary bone and The RTVs for different implant types and positions
were used for removal torque measurements. are shown in Figure 6. The pairwise Scheffé’s F tests
The removal torque testing resulted in a typical found a significant difference among the three implant
curve, where its peak was assumed to be the failure positions. The RTV were generally lower for more
(removal) torque of the bone–implant interface (Fig. posterior locations.
4). A summary of all removal torque values is shown The histologic analysis of the bone specimens fol-
in Table I for all healing periods and implant types. An lowing removal torque testing demonstrated a differ-
obvious difference was observed between the ma- ence between the smooth and the two rough surfaces.
chined surface and both the TPS and SLA surfaces. At For the machined implants, a separation between the
all time periods, the machined implants had mean bone and the implant surface was observed in all
RTVs <0.3 Nm, while the other implants demon- specimens [Fig. 7(a)]. In contrast, the two rough im-
strated mean RTVs >1.1 Nm. There was a significant plant surfaces often demonstrated fractures of bone
difference between the implant types (p < .00001) and trabeculae close to the implant surface, whereas the
implant positions (p < .0001), but not between the heal- bone–implant interface was still intact [Fig. 7(b)]. The
ing periods (p < .14). The mean RTVs for different
healing periods and implant types are shown in Fig-
ure 5. The pairwise Scheffé’s F tests found a significant
difference between the machined and the two rough

TABLE I
Removal Torque Values of Each Implant Type at
Different Healing Periods (in Newtonmeter)
4 wk 8 wk 12 wk
SLA TPS M SLA TPS M SLA TPS M

1.40 0.59 0.41 1.04 1.47 0.09 1.19 1.79 0.12

1.30 0.20 1.27 1.56 1.26 1.63 0.09
1.29 1.38 0.23 1.05 1.63 0.20 1.48 1.76 0.28
1.17 1.11 0.27 0.69 0.74 0.25 1.24 1.65 0.17
1.53 1.47 0.20 1.81 0.76 0.10 1.89 1.17
1.66 1.14 0.25 2.03 1.45 0.15 1.50 1.27 0.08
Mean 1.39 1.14 0.26 1.31 1.27 0.16 1.43 1.54 0.15
SD 0.18 0.34 0.08 0.51 0.41 0.07 0.26 0.26 0.08 Figure 6. Removal torque values of the three implant types
at different positions (anterior, middle, and posterior).
80 BUSER ET AL.

(a) (b)

Figure 7. (a) Histologic section of a machined implant: the separation occurred along the bone–implant interface (Toluidine
blue combined with basic fuchsin, original magnification ×12.5). (b) Histologic section of a TPS implant; the removal torque
test resulted in a fracture of a bone trabeculae close to the implant surface (black arrow). The bone–implant interface is still
intact (white arrows; Toluidine blue combined with basic fuchsin, original magnification ×25).

histomorphometric analysis revealed that the bone tested in the maxilla of miniature pigs by measuring
density was decreasing toward posterior implant lo- the interface shear strength, and were compared with
cations (Fig. 8) (p = .015). There was also a significant the two best documented titanium surfaces in implant
correlation between the local bone density and the dentistry, the machined and the titanium plasma-
measured removal torques (p < .005). sprayed (TPS) surface. This animal model was chosen
to measure removal torques of implants in maxillary
DISCUSSION bone because the bone structure of miniature pigs is
comparable to the bone structure of humans where
In the present study, titanium implants with a sand- dental implants are inserted in daily practice. For the
blasted and acid-etched surface were biomechanically implant shape, a screw design was chosen, since
threaded implants are generally preferred in implant
dentistry today. This meant that the interface shear
strength had to be evaluated by the measurement of
removal torques rather than pull- or push-out tests.
The tested solid-screw implants were identical in
shape and contained no macroscopic retentive ele-
ments such as vents in the apical portion or grooves in
the long axis. Based on the features selected, it was
assumed that any measured differences among the
three different implant types could most likely be at-
tributed to differences in their surface characteristics.
The characterization of the surface topography was
Figure 8. Bone density in implant sites: The density de- performed with stylus profilometry. This method is
creases toward posterior sites. well established and has been frequently used to char-
REMOVAL TORQUE VALUES OF SLA IMPLANTS 81

acterize titanium surfaces with different roughness iature pigs. Despite the favorable bone density in long
values.21 The analysis demonstrated that the TPS sur- bones of rabbits, the RTV in the rabbit studies never
face was the roughest surface among the three tested reached the level of 1.0 Nm with a machined surface.
surfaces, followed by the SLA surface, whereas the Some of these studies demonstrated that the anchor-
machined surface was clearly the smoothest surface. It age of titanium implants with a machined surface is
is important to note, however, that stylus profilometry clearly time dependent, since Gotfredsen et al.26 evalu-
cannot resolve the finer features of both, the TPS (un- ated a mean RTV of 0.32 Nm at 12 weeks of healing. In
dercuts) and the SLA surface (micropits). A more pre- contrast, Johansson and Albrektsson34 and Wenner-
cise method needs to be developed which is able to berg et al.29 observed mean RTVs of 0.88 and 0.62 Nm,
capture the small range features (1–10 ␮m) as well as respectively, after 12 months of healing in the tibia of
the long range features (10–30 ␮m). However, the rabbits. This time-dependent anchorage pattern for
measured values do represent the major topographical machined titanium surfaces underscores the impor-
features of each surface. tance of extended healing periods (6–8 months) for
The study confirmed the findings of several earlier this type of titanium surface in sites with poor bone
biomechanical studies in various animal models density or quantity.
which found that the surface texture of titanium im- The TPS surface was first described by Hahn and
plants has a significant influence on their anchorage in Palich36 and has been used in implant dentistry for
bone.17,18,22–30 The observed differences in this study more than 20 years. Retrospective and prospective
between the machined and the two rough (TPS and long-term studies with up to 20 years of follow-up
SLA) surfaces were highly significant (p < .00001). The have indicated excellent properties for successful
machined surface demonstrated mean removal torque long-term function.7,13,37 The TPS surface, however, is
values (RTV) between 0.26 and 0.15 Nm that were a coated surface, and disadvantages have also been
below the mean insertion torques, and showed a trend reported, such as the detachment of fine titanium par-
for decrease during the course of the study period. ticles from the implant surface during insertion, lead-
This decrease can be explained by the absence of mac- ing to titanium particles being located in the periim-
roscopic retention elements such as vents or grooves plant bone structure or even in regional lymph
in the long axis, where the ingrowth of newly formed nodes.38 Therefore, numerous alternative techniques
bone and subsequent bone remodeling could improve have been evaluated in recent years in an attempt to
implant anchorage over time. Therefore, a machined produce roughened titanium surfaces which do not
titanium surface does not provide a strong implant involve a coating procedure but are as favorable as the
anchorage in bone—particularly in compromised sites TPS surface with regard to long-term implant anchor-
with poor bone density or reduced vertical bone age in bone.
height, such as the posterior maxilla or the atrophic One of these techniques is based on a sandblasting
edentulous maxilla. This could explain significantly and acid-etching procedure. This technique has been
increased failure rates of titanium screw-type implants used to produce the SLA surface. This surface is not
with a machined surface in these compromised sites as microporous and has a lower roughness value Ra com-
reported by several authors.31–34 A difference between pared with the TPS surface (2.0 vs. 3.1 ␮m). It is char-
the tested surfaces was also apparent histologically. acterized by fine 1-–2-␮m micropits produced by a
The machined implants always demonstrated a sepa- strong acid bath with HCl/H2SO4 superimposed on
ration along the implant surface at the bone–implant the rough-blasted titanium surface. This new titanium
interface. The TPS and SLA surfaces, on the other surface has been intensively tested in recent years both
hand, often had fractures of bone trabeculae close to in vitro39–44 and in vivo.15,17–19,45 The in vivo studies
the implant surface, but an intact bone–implant inter- demonstrated predominantly superior results for the
face, indicating a strong physical interlock between SLA surface concerning implant integration and im-
the roughened titanium surface and bone. plant anchorage compared with the TPS surface, in
The RTV measured for machined implants in this particular during the initial healing period after im-
study are comparable to results of other removal plant placement. In the present study, the direct com-
torque studies in animals which tested titanium im- parison demonstrated higher mean RTVs for the SLA
plants with a machined surface,24–30,35 although the surface at 4 weeks of healing (1.39 vs. 1.14 Nm) with-
levels of the results in the present study were gener- out reaching statistical significance, owing to a rela-
ally slightly lower. This minor difference can most tively high standard deviation for both implant
likely be attributed to the fact that the other studies groups. The variation of RTV was primarily due to the
used a different animal model, since they were all car- local bone density, since the position of the implants
ried out in the tibia and/or femur of rabbits. The bone had a significant influence on the measured RTV.
structure in the rabbit tibia is characterized by a thick When interpreting these results, however, one should
cortex and has a much denser bone structure com- also be aware of the relatively small numbers of tested
pared with the alveolar process in the maxilla of min- implants within the three groups per time period and
82 BUSER ET AL.

the assumption of normal distribution in the statistical 2. Brånemark PI, Hansson BO, Adell R, Breine U, Lindström J,
analysis. At 8 and 12 weeks of healing, the mean RTVs Hallen O, Öhman A. Osseointegrated implants in the treat-
ment of the edentulous jaw: Experience from a 10-year period.
for both rough surfaces were similar. This trend for a Scand J Plast Reconstr Surg 1977;II(Suppl)16.
better and faster bone integration of SLA implants in 3. Schroeder A, Pohler O, Sutter F. Gewebereaktion auf ein Titan-
the initial healing period, reported by several authors, Hohlzylinderimplantat mit Titan-Spritzschichtoberfläche. Sch-
may be due to a higher production of local cytokines weiz Mschr Zahnheilk 1976;86:713–727.
and growth factors, as demonstrated by Kieswetter et 4. Schroeder A, Stich H, Straumann F, Sutter F. Über die Anla-
al.43 in an in vitro study with osteoblast-like cells. gerung von Osteozement an einen belasteten Implantatkörper.
Schweiz Mschr Zahnheilk 1978;88:1051–1058.
Recently, another acid-etched titanium surface was 5. Schroeder A, van der Zypen E, Stich H, Sutter F. The reaction
tested in an in vivo study. This study by Klokkevold et of bone, connective tissue and epithelium to endosteal im-
al.30 evaluated 4-mm screw-type titanium implants plants with sprayed titanium surfaces. J Maxillofac Surg 1981;
with an acid-etched surface (HCl/H2SO4) in the femur 9:15–25.
of rabbits and compared them to titanium implants 6. Adell R, Lekholm U, Rockler B, Brånemark PI. A 15-year study
of osseointegrated implants in the treatment of the edentulous
with the same macroscopic shape but a machined sur-
jaw. Int J Oral Surg 1981;10:387–416.
face. After 2 months of healing, the acid-etched sur- 7. Babbush CA, Kent JN, Misiek DJ. Titanium plasma-sprayed
faces demonstrated a four times higher resistance to (TPS) screw implants for the reconstruction of the edentulous
reverse torque rotation with a mean RTV at 0.20 ver- mandible. J Oral Maxillofac Surg 1986;44:274–282.
sus 0.05 Nm for machined surfaces, and confirmed a 8. Adell R, Eriksson B, Lekholm U, Brånemark PI, Jemt T. A
better bone anchorage of rough titanium surfaces. This long-term follow-up study of osseointegrated implants in the
treatment of totally edentulous jaws. Int J Oral Maxillofac Im-
acid-etched surface, however, is different from the
plants 1990;5:347–359.
SLA surface tested in the present study, since SLA 9. Buser D, Weber HP, Lang NP. Tissue integration of non-
implants are first sandblasted with a large grit (0.25– submerged implants: 1-year results of a longitudinal study
0.50 mm) prior to the acid-etching procedure. There- with ITI hollow-screw and hollow-cylinder implants. Clin Oral
fore, a direct comparison between these two surfaces Implant Res 1990;1:78–85.
might be of interest to evaluate the effect of the sand- 10. Buser D, Weber HP, Brägger U, Balsiger C. Tissue integration
of one-stage ITI implants: 3-year results of a longitudinal study
blasting procedure. A biomechanical study in minia- with hollow-cylinder and hollow-screw implants. Int J Oral
ture pigs is currently in progress to allow a direct Maxillofac Implants 1991;6:405–412.
comparison between the sandblasted and acid-etched 11. Lekholm U, van Steenberghe D, Herrmann I, Bolender C, Fol-
SLA surface versus this acid-etched alone surface. mer T, Gunne J, Henry P, Higuchi K, Laney W, Lindén U.
Osseointegrated implants in the treatment of partially edentu-
lous jaws: A prospective 5-year multicenter study. Int J Oral
CONCLUSIONS Maxillofac Implants 1994;9:627–635.
12. Mericske-Stern R, Steinlin-Schaffner T, Marti P, Geering AH.
The present study confirmed that surface character- Peri-implant mucosal aspects of ITI implants supporting over-
istics have an important influence on bone integration dentures: A 5-year longitudinal study. Clin Oral Implant Res
of titanium implants. The measurement of RTVs was 1994;5:9–18.
used to evaluate the interface shear strength of tita- 13. Buser D, Mericske-Stern R, Bernard JP, Behneke A, Behneke N,
nium implants with three different surfaces in the Hirt HP, Belser UC, Lang NP. Long-term evaluation of non-
submerged ITI implants. Part I: 8-year life table analysis of a
maxilla of miniature pigs. The machined titanium sur- prospective multi-center study with 2359 implants. Clin Oral
face clearly demonstrated the lowest mean removal Implant Res 1997;8:161–172.
torques (0.15–0.25 Nm), whereas the two rough sur- 14. Thomas KA, Cook S. An evaluation of variables influencing
faces—titanium plasma-sprayed (TPS) and sand- implant fixation by direct bone apposition. J Biomed Mater Res
blasted, acid-etched (SLA)—exhibited both mean val- 1985;19: 875–901.
15. Buser D, Schenk RK, Steinemann S, Fiorellini JP, Fox CH, Stich
ues clearly exceeding 1.14 Nm (1.14–1.54 Nm). There
H. Influence of surface characteristics on bone integration of
were no significant differences between the TPS and titanium implants: A histometric study in miniature pigs. J
SLA surfaces, although SLA implants demonstrated Biomed Mater Res 1991;25:889–902.
slightly higher mean RTVs at 4 weeks of healing. 16. Steinemann S. The properties of titanium. In: Schroeder A,
Sutter F, Buser D, Krekeler G, editors. Oral implantology: Ba-
The authors are grateful to Dr. D. Mettler, V.D., and the sics, ITI Dental Implant System. 2nd ed. New York: Thieme
whole team of the Surgical Research Unit ESI and Clinic for Medical; 1996. p 37–59.
Large Animals, University of Berne, for their excellent sup- 17. Wilke HJ, Claes L, Steinemann S. The influence of various ti-
port during the study. The help of C. Grau during surgeries tanium surfaces on the interface shear strength between im-
is highly appreciated. The authors also thank B. Hoffmann plants and bone. In: Heimke G, Soltész U, Lee AJC, editors.
and D. Reist for the histologic and histometric analysis. Advances in biomaterials, vol 9: Clinical implant materials.
Amsterdam: Elsevier Science Publishers BV; 1990. p 309–314.
18. Wong M, Eulenberger J, Schenk RK, Hunziker E. Effect of sur-
References face topography on the osseointegration of implant materials
1. Brånemark PI, Breine U, Adell R, Hansson BO, Lindström J, in trabecular bone. J Biomed Mater Res 1995;29:1567–1575.
Ohlsson å. Intra-osseous anchorage of dental prostheses. I. Ex- 19. Cochran DL, Schenk RK, Lussi A, Higginbottom FL, Buser D.
perimental studies. Scand J Plast Reconstr Surg 1969;3:81–100. Bone response to unloaded and loaded titanium implants with
REMOVAL TORQUE VALUES OF SLA IMPLANTS 83

a sandblasted and acid-etched surface: A histometric study in A three-year follow-up study in 70 patients. Clin Oral Implant
the canine mandible. J Biomed Mater Res 1998;40:1–11. Res 1993;4:187–194.
20. Schenk RK, Olah AJ, Herrmann W. Preparation of calcified 34. Jemt T, Chai J, Harnett J, Heath MR, Hutton JE, Johns RB,
tissue for light microscopy. In: Dickson, GR, editor. Methods of McKenna S, McNamara DC, van Steenberghe D, Taylor R,
calcified tissue preparation. Amsterdam: Elsevier Science Pub- Watson RM, Hermann I. A 5-year prospective multicenter fol-
lishers BV; 1984. p 1–56. low-up report on overdentures supported by osseointegrated
21. Wennerberg A, Albrektsson T, Ulrich H, Krol J. An optical 3D implants. Int J Oral Maxillofac Implants 1996;11:291–298.
method for topographical description of surgical implants. J 35. Johansson C, Albrektsson T. Integration of screw implants in
Biomed Eng 1992;14:412–418. the rabbit: A 1-year follow-up of removal torque of titanium
22. Claes L, Hutzschenreuter P, Pohler O. Lösemomente von Cor- implants. Int J Oral Maxillofac Implants 1987;2:69–75.
ticalisschrauben in Abhängigkeit von Implantationszeit und 36. Hahn H, Palich W. Preliminary evaluation of porous metal
Oberflächenbeschaffenheit. Arch Orthop Unfall-Chir 1976;85: surfaced titanium for orthopedic implants. J Biomed Mater Res
155–159. 1970;4:571–579.
23. Thomas KA, Kay JF, Cook SD, Jarcho M. The effect of surface 37. Ledermann PD. Über 20jährige Erfahrung mit der sofortigen
macrotexture and hydroxylapatite coating on the mechanical funktionellen Belastung von Implantatstegen in der regio in-
strengths and histologic profils of titanium implant materials. terforaminalis. Z Zahnärztl Implantol 1996;12:123–136.
J Biomed Mater Res 1987;21:1395–1414. 38. Weingart D, Steinemann S, Schilli W, Strub JR, Hellerich U,
24. Carlsson L, Röstlund T, Albrektsson B, Albrektsson T. Re- Assenmacher J, Simpson J. Titanium deposition in regional
moval torques for polished and rough titanium implants. Int J lymph nodes after insertion of titanium screw implants in the
Oral Maxillofac Implants 1988;3:21–24. maxillofacial region. Int J Oral Maxillofac Surg 1994;23:450–
25. Gotfredsen K, Nimb L, Hjørting-Hansen E, Jensen JS, Holmen 452.
A. Histomorphometric and removal torque analysis for 39. Cochran DL, Simpson J, Weber HP, Buser D. Attachment and
smooth and TiO2-blasted titanium implants in dogs. Clin Oral growth of periodontal cells on smooth and rough titanium. Int
Implant Res 1992;3:77–84. J Oral Maxillofac Implants 1994;9:289–297.
26. Gotfredsen K, Wennerberg A, Johansson C, Skovgaard LT, 40. Martin JY, Schwartz Z, Hummert TW, Schraub DM, Simpson J,
Hjørting-Hansen E. Anchorage of TiO2-blasted, HA-coated, Lankford J Jr, Dean D, Cochran DL, Boyan BD. Effect of tita-
and machined implants: An experimental study with rabbits. J nium surface roughness on proliferation, differentiation, and
Biomed Mater Res 1995;29:1223–1231. protein synthesis of human osteoblast-like cells (MG63). J
27. Wennerberg A, Albrektsson T, Andersson B, Krol JJ. A histo- Biomed Mater Res 1995;29:389–401.
metric and removal torque study on screw-shaped titanium 41. Martin JY, Dean DD, Cochran DL, Simpson J, Boyan BD,
implants with three different surface topographies. Clin Oral Schwartz Z. Proliferation, differentiation, and protein synthe-
Implant Res 1995;6:24–30. sis of human osteoblast-like cells (MG63) cultured on previ-
28. Wennerberg A, Albrektsson T, Lausmaa J. A torque and his- ously exposed titanium surfaces. Clin Oral Implant Res 1996;
tomorphometric evaluation of c.p. titanium screws, blasted 7:27–37.
with 25 and 75 ␮m sized particles of Al2O3. J Biomed Mater Res 42. Schwartz Z, Martin JY, Dean DD, Simpson J, Cochran DL,
1996;30:251–260. Boyan BD. Effect of titanium surface roughness on chrondo-
29. Wennerberg A, Ektessabi A, Albrektsson T, Johansson C, cyte proliferation, matrix production, and differentiation de-
Andersson B. A 1-year follow-up of implants of differing sur- pends on the state of cell maturation. J Biomed Mater Res
face roughness placed in rabbit bone. Int J Oral Maxillofac 1996;30:145–155.
Implants 1997;12:486–494. 43. Kieswetter K, Schwartz Z, Hummert TW, Cochran DL, Simp-
30. Klokkevold PR, Nishimura RD, Adachi M, Caputo A. Osseo- son J, Dean DD, Boyan BD. Surface roughness modulates the
integration enhanced by chemical etching of the titanium sur- local production of growth factors and cytokines by osteoblast-
face: A torque removal study in the rabbit. Clin Oral Implant like MG-63 cells. J Biomed Mater Res 1996;32:55–63.
Res 1997;8:442–447. 44. Boyan BD, Batzer R, Kieswetter K, Liu Y, Cochran DL, Szmuck-
31. Jaffin RA, Berman CL. The excessive loss of Brånemark fixtures ler-Moncler S, Dean DD, Schwartz Z. Titanium surface rough-
in type IV bone: A 5-year analysis. J Periodontol 1991;62:2–4. ness alters responsiveness of MG63 osteoblast-like cells to
32. Quirynen M, Naert I, van Steenberghe D, Schepers E, Calber- 1␣,25-(OH)2D3. J Biomed Mater Res 1998;39:77–85.
son L, Theuniers G, Ghyselen J, de Mars G. The cumulative 45. Cochran DL, Nummikowski PV, Higginbottom FL, Hermann
failure rate of the Brånemark system in the overdenture, the JS, Makins SR, Buser D. Evaluation of an endosseous titanium
fixed partial, and the fixed full prostheses design: A prospec- implant with a sandblasted, acid-etched surface in the canine
tive study on 1273 fixtures. J Head Neck Pathol 1991;10:43–53. mandible: Radiographic results. Clin Oral Implant Res 1996;7:
33. Jemt T. Implant treatment in resorbed edentulous upper jaws: 240–252.