Influence of surface characteristics on bone

integration of titanium implants.

A histomorphometric study in miniature pigs

D. Buser
Department of Oral Surgery, School of Dental Medicine, University of Berne, Switzerland

R. K. Schenk
Department of Pathophysiology, University of Berne, Switzerland

S. Steinemann
Department of Experimental Physics, University of Lausanne, Switzerland

J. P. Fiorellini
Department of Periodontology, Harvard School of Dental Medicine, Boston, Massachussets

C. H. Fox
Department of Dental Care Administration, and Department of Periodontology,
Harvard School of Dental Medicine, Boston, Massachussets

H. Stich
Department of Operative Dentistry, School of Dental Medicine, University of Berne,
Switzerland

The purpose of the present study was to (medium grit; HF/HN03) implant sur-
evaluate the influence of different surface faces had the lowest percentage of bone
characteristics on bone integration of tita- contact with mean values ranging be-
nium implants. Hollow-cylinder implants tween 20 and 25%. Sandblasted implants
with six different surfaces were placed in with a large grit and titanium plasma-
the metaphyses of the tibia and femur sprayed implants demonstrated 30-40%
in six miniature pigs. After 3 and 6 weeks, mean bone contact. The highest extent of
the implants with surrounding bone were bone-implant interface was observed in
removed and analyzed in undecalcified sandblasted and acid attacked surfaces
transverse sections. The histologic ex- (large grit; HC1/H2S04)with mean val-
amination revealed direct bone-implant ues of 50-60%, and hydroxylapatite (HA)-
contact for all implants. However, the coated implants with 60-70%. However,
morphometric analyses demonstrated sig- the HA coating consistently revealed
nificant differences in the percentage of signs of resorption. It can be concluded
bone-implant contact, when measured in that the extent of bone-implant interface
cancellous bone. Electropolished as well is positively correlated with an increas-
as the sandblasted and acid pickled ing roughness of the implant surface.

Journal of Biomedical Materials Research, Vol. 25, 889-902 (1991)

0 1991 John Wiley & Sons, Inc. CCC 0021-9304/91/070889-14$4.00
890 BUSER ET AL.

IN TRODUCTION

Replacement of missing teeth in partially or fully edentulous patients has

most often involved the use of fixed or removable prostheses supported by
natural teeth and/or soft tissues. In the past 30 years, the use of dental im-
plants has become a viable treatment alternative for these patients. Subpe-
riosteal implants, commonly used in the 1950s, were replaced in importance
in the late 1960s by endosseous implants. These implants are placed with ini-
tial stability using a precise fit in the jaw bone, which requires no cement for
stabilization. Fundamental studies, conducted by Brinemark et al.Ir2and
Schroeder et al.3-5using commercially pure (cp) titanium implants in animal
models, showed that implant anchorage with direct bone contact can be
achieved. This direct bone-implant interface was termed osseointegration.'
In the past 10 years, these findings have been confirmed by several research
Subsequently, successful osseointegration was also reported in
clinical studies using cp titanium irnplant~.~-''
The percentage of bone-implant contact necessary to create sufficient an-
chorage to permit successful implant function as a load-bearing device over
time remains unclear. Some clinical studies have demonstrated that areas of
the jaws with extremely cancellous bone, such as the posterior maxilla, offer
significantly lower success rates when compared with areas of denser struc-
ture.9,1~,i3 Therefore, several attempts have been made in the past 20 years to
improve implant anchorage through the use of porous implant surfaces or
different coating materials. Orthopedic implants with varying porous sur-
faces have been studied widely since the late 1960~.'~,'~ Plasma-spray coating
of implants to create a porous surface was first reported by Hahn and Palich
in 1970.16 Experimental studies using implants with a titanium plasma-
sprayed (TI'S) surface demonstrated significantly higher removal t o r q ~ e s , ' ~ , ' ~
and a higher percentage of direct bone-implant contact19 when compared
with smooth titanium implants. Similarly, the International Team for Oral
Implantology (ITI) has successfully used implants with a TI'S surface since
1974.'" In recent years, the plasma-spray technique has also been used to coat
titanium implants with hydroxylapatite (HA). Experimental studies showed a
higher percentage of bone-implant contact and greater values of maximum
shear strength for HA-coated implants when compared with uncoated or
grit-blasted titanium An alternative technique to produce rough
titanium surfaces is sandblasting. This technique can be used with or without
acid treatment. Recently, various titanium surfaces using this technique were
tested by Wilke et a1.28in a biomechanical study measuring removal torques
of cortical screws.
The purpose of the present study was the examination of bone reactions to
titanium implants with various surface characteristics to extend the bio-
mechanical results reported by Wilke et a1." Smooth and TPS implant surfaces
were compared to implant surfaces produced by alternative techniques such as
sandblasting, sandblasting combined with acid treatment, and plasma-coating
with HA. The evaluation was performed with histomorphometric analyses
measuring the extent of the bone-implant interface in cancellous bone.
BONE INTEGRATION OF TITANIUM IMPLANTS 891

MATERIALS A N D METHODS

Implant design and surfaces

Commercially pure titanium implants with six different surfaces were

placed in both the metaphyses of the distal femur and tibia1 heads in six
12-month-old miniature pigs. All implants were perforated hollow-cylinders
with outer diameters of 3.55 mm and lengths of 10 mm (Fig. 1).The six differ-
ent surface characteristics studied were:
Type E = electro-polished
Type SMP = sandblasted with medium grit (0.12-0.25 pm) and acid
pickling with HF/HN03
Type SL = sandblasted with large grit (0.25-0.50 pm)
Type SLA = sandblasted with large grit (0.25-0.50 pm) and acid
attack with HC1/H2S04
Type TI'S = Titanium plasma-sprayed
Type HA = HA plasma-sprayed
The above treatments produced the following surface characteristics. The
E surface was flat and smooth. The SMP surface (Fig. 2A) showed structure,
deformation markings and crystal line characteristics; the average roughness
being 6 pm. The SL surface (Fig. 2B) had roughness values of 22-28 pm.
Corundum particles were embedded in the highly distorted metal. Etching
in a reducing acid mixture (type SLA; Fig. 2C) resulted in a rough surface
(roughness 18-23 pm) with an additional fine secondary structure. The TPS
coating (Fig. 2D) had typical roughness values of 30-50 pm and an apparent
porosity. The structure of the HA coating (original IMZ coating; Dr. A.
Kirsch, Filderstadt, Germany) was comparable to the TPS surface (Fig. 2E).

Surgical procedure

A total of 72 implants, 12 of each type, were placed under general anesthe-

sia in aseptic surgical conditions (Surgical Research Unit ESI and Clinic for

Figure 1. All experimental implants had the same basic shape of perfo-
rated hollow-cylinders, but six different surfaces. From left to right: type
TPS, SMP, SL, SLA, E, and HA.
892 BUSER ET AL.

(El
Figure 2. Scanning electron micrographs of different implant surfaces
(bar is 20 pm; original magnification ~ 1 7 0 0 ) (A)
. type SMP, (B) type SL,
(C) type SLA, (D) type TPS, (E) type HA.

Large Animals, University of Berne). Prior to the surgery, the animals were
given 1 g of prophylactic amoxicillin IM. The recipient sites in the metaphy-
ses were prepared using original IT1 instruments (Institut Straumann AG,
Waldenburg, Switzerland) with an outer diameter of 3.50 mm. Thus, primary
stability was achieved by the press-fit design. Bone preparation was per-
formed with rotary instruments at 500 rpm using copious irrigation with
chilled sterile physiological saline. Three implants of the same type were
inserted per surgical site into the tibia or femur. Four different implant
types were tested in each animal. All implants exhibited primary stability.
Wound closure of different soft tissue layers was achieved with resorbable
suture material.
BONE INTEGRATION OF TITANIUM IMPLANTS 893

Histologic preparation

The miniature pigs were sacrificed at either 3 or 6 weeks following

surgery. Block sections containing the implants were fixed in 10% buffered
formaldehyde. A radiograph was made of each specimen. The specimens
were dehydrated and embedded in methylmethacrylate. Undecalcified sec-
tions of -500 pm in thickness were obtained using a low-speed diamond
saw with coolant. Approximately eight transverse sections of each implant
could be obtained. Subsequently, four sections of each implant were ground
to a final thickness of -80 pm and stained with toluidine blue.29
A preliminary review of the sections revealed that the implants most dis-
tant from the knee joint were partially located in the marrow space of the
long bones in some animals. Since the extent of bone-implant contact was to
be examined in cancellous bone, only the two implants placed closest to the
epiphysis were histomorphometrically analyzed resulting in a total number
of 48 evaluated implants, eight in each animal. One hundred ninety-two
transverse sections, four of each implant, were histologically prepared. Of
these, four specimens could not be used for the analyses due to technical arti-
facts. Thus, 188 sections were morphometrically evaluated. The quantita-
tively analyzed material represents a reference surface of -1050 mm’. The
percentage of direc: bone-implant contact was only evaluated in cancellous
bone at the outer circumference of the implants using a light microscope
(magnification ~ 1 0 0 ,10 x 10 grid) and the morphometric technique de-
scribed by Schenk and Olah.30

Statistical analysis

The mean percentage ( 5 standard deviation) of bone-implant contact for

each implant type was analyzed at 3 and 6 weeks. In addition, an analysis of
variance F test was used to test the null hypothesis of no difference between
the various implant groups with the following model:
X,k =p + a, + PI + AB, + E V k
where
Yllk= Percent bone contact
p = Grand mean
a, = Fixed effect for the ith implant
PI = Random effect for the jth pig
AB, = Random effect for the interaction
E,lk = Error term
To control the type I error rate, ScheffBs method for multiple comparisons
was used to investigate pairwise comparison. Analyses of variance and pair-
wise comparisons were computed with the GLM procedure of SAS release
6.03 (SAS Institute Inc., Cary, NC, USA).
894 BUSER ET AL.

RESULTS

The surgical sites showed no evidence of infection throughout the healing

periods and up to the date of sacrifice. All implants revealed clinically firm
anchorage in bone (Fig. 3) and the radiographs indicated no signs of
peri-implant radiolucencies (Fig. 4). All implant types demonstrated histolog-
ically direct bone contact. The morphometric analyses, however, yielded dif-
ferences in the extent of bone-implant interface among the six implant types.
The morphometric results are summarized in Table I, and graphically de-
picted in Table 11. The lowest percentages of direct bone contact were ob-
served for E and SMP implants with mean values ranging between 20 and
25% after 3 and 6 weeks (Figs. 5A,B). The corresponding values for SL im-
plants were slightly higher with means of 30-35% (Fig. 5C). A further in-
crease of bone contact was demonstrated with TI'S implants with means
slightly below 40% at both observation periods (Fig. 5D). SLA implants re-
sulted in significantly higher mean values ranging between 50 and 60%.
Around SLA implants, a narrow band of bone was often apparent as an ex-
tension of perpendicularly oriented bone trabeculae (Fig. 5E). These findings
were also observed around HA implants, which demonstrated the highest ex-
tent of bone-implant interface with means of 60-70% (Fig. 5F). However, the
HA coating consistently revealed signs of resorption in both, the 3- and

Figure 3. Block section from a tibia head containing three implants.

Figure 4. Radiograph of the block section in Figure 3 showing no signs of

peri-implant radiolucencies.
BONE INTEGRATION OF TITANIUM IMPLANTS 895

TABLE I
Percentages of Direct Bone-Implant Contact at 3 and 6 Weeks
for the Six Different Implant Surfaces
Period TY Pe n Min Max Mean SD
3 weeks E 16 7.3 54.3 24.9 14
SMP 16 2.6 41.4 22.3 10.6
SL 16 9.1 55.2 30.4 11.6
T PS 14 26.2 60.7 39.2 12
SLA 16 35.3 86.8 52.1 13
HA 14 31.6 76.3 60.6 11.6
6 weeks E 16 13.1 41.4 25.1 7.4
SMP 16 5.7 41.2 21.6 9.3
SL 16 17.1 62.5 33.6 11.5
TI’S 16 21.1 64.7 37.8 11.6
SLA 16 41.6 74.1 57.7 9.5
HA 16 57.3 77.9 69.5 6.5
n = Number of evaluated histologic sections.

6-week specimens. These sites of resorption were observed predominantly in

areas without bone covering, where marrow cells had access to the apatite
coating (Fig. 6).
The analyses of variance in Table I11 and Table IV for the mean percentage
of direct bone-implant contact, show the overall F test to be significant
for the model specified in the materials and methods section for both obser-
vation periods. The partial F tests for both the pig effect and the implant ef-
fect were significant at the 0.05 level at 3 weeks, whereas only the implant

TABLE I1
Mean Percentages of Direct Bone Contact of Different
Implant Surfaces at 3 and 6 Weeks
896 BUSER ET AL.

(El (F)
Figure 5. Photomicrographs of the six different implant surfaces (tolu-
idine blue; original magnification X25). (A) type E implant, (B) type SMP
implant, (C) type SL implant, (D) type TPS implant, (E) type SLA implant,
(F) type HA implant.

effect was significant at 6 weeks. Controlling the type I error rate, Scheffgs
method was used for multiple pairwise comparisons of the mean percentage
of direct bone-implant contact between the different implant types. Overall,
HA implants revealed a significantly greater percentage of bone-implant
contact than E, SMP, SL, and TPS implants at both observation periods. SLA
implants showed a significant difference to E, SMP, and SL implants, whereas
TPS implants resulted in statistically higher values than E, and SMP implants
at both observation periods. Corresponding p-values are listed in Table V and
Table VI.
BONE INTEGRATION OF TITANIUM IMPLANTS 897

Figure 6. Photomicrograph showing signs of resorption of the HA coating

(*). These resorptions (arrows) were observed predominantly in areas
where marrow cells had access to the apatite coat (UV-light; toluidine and
rhotamin-staining; original magnification ~ 2 0 0 ) .

DISCUSSION

In the present study, the bone reaction to titanium implants with six differ-
ent surfaces was morphometrically analyzed in miniature pigs. To control
the influence of different variables, the surgical procedure and the basic di-
mensions of all six implant types were standardized. Only the influence of
the pig, the influence of the implant surface, and random error were uncon-
trolled. Therefore, other than random error, observed differences in the mean
'percentage of 'bone-implant contact can be attributed mainly to the pig
and/or the implant effect. Three animals were sacrificed after 3 weeks, a
TABLE I11
The Percentage of Bone Contact at 3 Weeks ANOVA Table for the Model Specified
and for Each Effect in the Model
Source DF Sum of Squares Mean Square F Value Pr > F
Model 11 20821.3 1892.8 15.29 0.0001
Error 80 9904.9 123.8
Total 91 30726.2

Source DF Type I11 SS Mean Square F Value Pr > F

Implant 5 14826.5 2965.3 11.70 0.0168
Pig 2 1701.4 850.7 6.87 0.0018
Interaction 4 1013.6 253.4 2.05 0.0957

Note: The denominator for the F value for the pig and pig-implant interaction is the
mean square error; while for the implant effect, it is the mean square of the interaction.
898 BUSER ET AL.

TABLE IV
The Percentage of Bone Contact at 6 Weeks ANOVA Table for the Model Specified
and for Each Effect in the Model
Source DF Sum of Squares Mean Square F Value Pr > F
Model 11 29412.1 2673.8 31.01 0.0001
Error 84 7243.5 86.2
Total 95 36655.6

Source DF Type I11 SS Mean Square F Value Pr > F

Implant 5 24345.8 4869.2 35.80 0.002
Pig 2 359.6 179.8 2.09 0.13
Interaction 4 544.0 136.0 1.58 0.19

Note: The denominator for the F value for the pig and pig-implant interaction is the
mean square error; while for the implant effect, it is the mean square of the interaction.

TABLE V
Results of ScheffC's Method for Multiple Pairwise Comparisons in an ANOVA:
Percentage of Bone Contact at 3 weeks
E SMP SL TPS SLA HA
~~ ~

E - N.S. N.S. 0.05 0.001 0.001

SMP - N.S. 0.01 0.001 0.001
SL - N.S. 0.001 0.001
T PS - N.S. 0.001
SLA - N.S.
HA -

N.S. = no statistically significant difference at p < 0.05.

TABLE VI
Results of ScheffC's Method for Multiple Pairwise Comparisons in an ANOVA:
Percentage of Bone Contact at 6 Weeks
E SMP SL TPS SLA HA
E - N.S. N.S. 0.05 0.001 0.001
SMP - 0.05 0.001 0.001 0.001
SL - N.S. 0.001 0.001
TI'S - 0.001 0.001
SLA - 0.05
HA -
~ ~~ ~ ~ ~~

N.S.= no statistically significant difference at p < 0.05.

second group of three animals i-IlLer b weeks. These intervals were chosen
based on previous studies of gap healing during fracture repair in dogs.31A
comparison of the present results indicates that the activation of bone forma-
tion in miniature pigs is extremely rapid, since the bone-implant interface is
nearly as high after 3 weeks as after 6 weeks. The morphometric analyses re-
vealed a clear tendency for an increase of bone-implant contact with an in-
creasing roughness of the implant surface. This confirms earlier observations
in a morphometric study reported by Strunz et al.I9This may also explain re-
BONE INTEGRATION OF TITANIUM IMPLANTS 899

sults in biomechanical studies demonstrating consistently higher removal

torques for TPS implants when compared with smooth implant^.'^,'^ The same
tendency was observed when the three different sandblasted surfaces were
compared. The least rough SMP surface demonstrated the lowest percentage
of bone contact within this group. The direct comparison of SL and SLA sur-
faces, both sandblasted with the same grit size, revealed a significant differ-
ence with an increase to almost 60% of bone contact for SLA implants. The
difference can most likely be attributed to the acid treatment with HC1 and
HzS04.This particular acid mixture not only removed remaining grit parti-
cles embedded in the titanium surface, but also attacked the titanium surface
producing a secondary roughness which is apparent in scanning electron mi-
crographs (Fig. 2C). However, it is not clear yet if the acid bath could have
altered other characteristics, such as electrochemical properties of the im-
plant surface, which might stimulate bone formation around implants. Four
of the six titanium surfaces were recently tested in a biomechanical study by
Wilke et a1.,28in which the removal torques of cortical screws were measured
in sheep tibia. The results of this 52-week study demonstrated the highest re-
moval torques for SLA surfaces to be 6.9 Nm at 24 weeks. Second highest val-
ues were measured for TPS surfaces at 5.3 Nm after the same period. In
contrast, E and SMP surfaces mostly resulted in removal torques below the
insertion torque of 1.0 Nm.
In the present study, HA-coated implants demonstrated the highest per-
centage of bone-implant contact. Therefore, the HA coating seemed to have a
stimulating influence on bone apposition within 6 weeks of healing. These
findings confirm results of other studies with HA coated implants.22,26 How-
ever, the HA coating used in this study consistently revealed signs of resorp-
tion. Thus, it may be concluded that this particular HA coating is biologically
unstable. In fact, x-ray diffraction of the HA coating gave a predominantly
amorphous pattern with a crystallinity below 207G.3’ This state of the calcium
phosphate is expected to be less stable and prone to biologic degradation.
These findings confirm observations in other studies where resorption of HA
plasma-sprayed coatings was also r e p ~ r t e d . ~ ” ~ ~

CONCLUSIONS

The morphometric results of the present study may be summarized as

follows:
(1) Rough implant surfaces generally demonstrated an increase in bone ap-
position compared to polished or fine structured surfaces.
(2) The acid treatment with HC1/H2S0, used for SLA implants had an ad-
ditional stimulating influence on bone apposition.
(3) The HA-coated implants showed the highest extent of bone-implant
interface.
(4) The HA coating used in this study (original IMZ coating) consistently
revealed signs of resorption.
900 BUSER ET AL.

Further studies are necessary to determine which surface characteristics

are optimal for implants placed in areas of cancellous bone. Based on the
present study, sandblasting and chemical etching with HC1/H2S04as well as
HA coating seem to be the most promising alternatives to titanium implants
with smooth or TI'S surfaces. Future studies should evaluate if HA coatings
are available which are biologically stable. In addition, it would be interesting
to compare SLA titanium surfaces directly with different HA coatings in me-
chanical push-out tests.

The authors thank Dr. Schilt, Dr. Mettler, and the Staff of the Surgical Research Unit
ESI and Clinic for Large Animals, University of Berne, for their assistance. In addition,
the help of B. Hoffmann, G. Leuenberger, D. Reist, Dr. T. H. Howell, and Dr. R.W. Tracy
is highly appreciated. The study was supported by the M. E. Miiller Foundation, Berne,
the ITI-Foundation, Waldenburg, and the Swiss Foundation for Medical and Biological
Stipends, Basel.

References
1. I? I. Brbnemark, U. Breine, R. Adell, et al, "Intra-osseous anchorage of
dental prostheses. I. Experimental studies," Scand. J. Plast. Xeconstr.
Surg., 3, 81-100, (1969).
2. I? 1. Brbnemark, B.O. Hansson, R. Adell, et al, "Osseointegrated im-
plants in the treatment of the edentulous jaw. Experience from a
10-year period," Scand. I. Plast. Reconstr. Surx., 11, suppl 16, (1977).
3. A. Schroeder, 0. Pohler, and F. Sutter, "Gewebereaktion auf ein Titan-
Hohlzylinderimplantat mit Titan-Spritzschichtoberflache," Schweiz.
Mschr. Zahnheilk., 86, 713-727 (1976).
4. A. Schroeder, H. Stich, F. Straumann, and F. Sutter, "Uber die An-
lagerung von Osteozement an einen belasteten Implantatkorper,"
Schweiz. Mschr. Zahnheilk., 88, 1051-1058 (1978).
5. A. Schroeder, E. van der Zypen, H. Stich, and F. Sutter, "The reaction
of bone, connective tissue and epithelium to endosteal implants with
sprayed titanium surfaces," I. Max. Fac. Surg., 9, 15-25 (1981).
6. A. Kirsch and K. Donath, "Tierexperimentelle Untersuchungen zur
Bedeutung der Mikromorphologie von Titanimplantatoberflachen,"
Fortschr. Zahnarztl. lmplantol., 1, 35-40 (1984).
7. D. A. Deporter, P. A. Watson, R. M. Pilliar, et al., 'A histological assess-
ment of the initial healing response adjacent to porous-surfaced, tita-
nium alloy dental implant system in dogs," J. Dent. Res., 65, 1064-1070
(1986).
8. E. Roberts, "Bone tissue interface," J. Dent. Educ., 52,804-809 (1988).
9. R. Adell, U. Lekholm, B. Rockler, and P. I. Brinemark, "A 15-year study
of osseointegrated implants in the treatment of the edentulous jaw,"
Int. J. Oral Surg., 10, 387-416 (1981).
10. C. A. Babbush, J. N. Kent, and D. J. Misiek, "Titanium plasma-sprayed
(TI'S) screw implants for the reconstruction of the edentulous
mandible," J. Oral Maxillofac. Surg., 44, 274-282 (1986).
11. T. Albrektsson, E. Dahl, L. Enbom, et al., "Osseointegrated oral im-
plants. A Swedish multicenter study of 8139 consecutively inserted
nobelpharma implants," J. Periodontol., 59, 287-296 (1988).
12. D. Buser, H.P. Weber, and N.P. Lang, "Tissue integration of non-
submerged implants. 1-year results of a longitudinal study with IT1
hollow-screw and hollow-cylinder implants," Clin. Oral Impf. Res., 1,
33-40 (1990).
BONE INTEGRATION OF TITANIUM IMPLANTS 901

13. R. A. Jaffin and C. L. Berman, "The excessive loss of Branemark Fix-

tures in Type IV Bone: A 5-year Analysis," J. Periodontol., 62,Z-4 (1991).
14. R. M. Pilliar, "Porous-surfaced metallic implants for orthopedic appli-
cations," J. Biomed. Muter. Res., 21, 1-33 (1987).
15. M. Spector, "Historical review of porous-coated implants," J. Arthropl.,
2, 163-177 (1987).
16. H. Hahn and W. Palich, "Preliminary evaluation of porous metal sur-
faced titanium for orthopedic implants," J. Biomed. Muter. Res., 4, 571-
577 (1970).
17. L. Claes, P. Hutzschenreuter, and 0. Pohler, "Losemomente von
Corticalisschrauben in Abhangigkeit von Implantationszeit und
Oberf Iachen-beschaffenheit," Arch. ovthop. Unfall-Chir., 85, 155-159
(1976).
18. S.G. Steinemann, J. Eulenberger, P. A. Mausli, and A. Schroeder, 'Adhe-
sion of bone to titanium," Advances in Biomaterials, Vol. 6: Biological and
Biomechanical Performances of Biomaterials, P. Christel, A. Meunier, and
A. J.C. Lee (eds.), Elsevier Science Publishers B.V., Amsterdam, 1986,
pp. 409-414.
19. V. Strunz, U. Gross, and S. Nickel, "Morphometrische Untersuchungen
iiber den Knochenkontakt an titanplasmabeschichteten Implantaten,"
Fortschr. Kiefer Gesichtsckir., 28, 47-49 (1983).
20. A. Schroeder, F. Sutter, and G. Krekeler, "Oral Implantology, " Georg
Thieme Verlag, Stuttgart, 1991.
21. K. de Groot, R. Geesink, C. P. A.T. Klein, and P. Serekian, "Plasma
sprayed coatings of hydroxylapatite," J. Biomed. Muter. Res., 21, 1375-
1381 (1987).
22. K.A. Thomas, J.F. Kay, S. D. Cook, and M. Jarcho, "The effect of
surface macrotexture and hydroxylapatite coating on the mechani-
cal strengths and histologic profils of titanium implant materials,"
J. Biomed. Muter. Res., 21, 1395-1414 (1987).
23. J. F. Osborn, "Die biologische Leistung der HA-Keramik-Beschichtung
auf dem Femurschaft-Erste histologische Auswertung eines Human-
explantates," Biomed. Tech., 32, 177-183 (1987).
24. M. S. Block, J. N. Kent, and J. F. Kay, "Evaluation of hydroxylapatite-
coated titanium dental implants in dogs," J. Oral Maxillofac. Surg., 45,
601-607 (1987).
25. G. L. de Lange and K. Donath, "Interface between bone tissue and im-
plants of solid hydroxyapatite or hydroxyapatite-coated titanium
implants," Biomaterials, 10, 121-125 (1989).
26. M.S. Block, I. M. Finger, M.G. Fontenot, and J. N. Kent, "Loaded
hydroxylapatite-coated and grit-blasted titanium implants in dogs,"
Int. J. Oral Maxillofac. Implants., 4, 219-226 (1989).
27. H.W. Denissen, W. Kalk, H. M. de Nieuport, J.C. Maltha, and A. van de
Hooff, "Mandibular bone response to plasma-sprayed coatings of
hydroxy-apatite," Int. J. Prosth., 3, 53-58 (1990).
28. H. J. Wilke, L. Claes, and S. Steinemann, "The influence of various ti-
tanium surfaces on the interface shear strength between implants and
bone," in Advances in Biomaterials, Vol. 9: Clinical Implant Materials, G.
Heimke, U. Solthsz, and A. J.C. Lee (eds.), Amsterdam, Elsevier Science
Publishers B.V., 1990, pp. 309-314.
29. R. K. Schenk, A. J. Olah, and W. Hermann, "Preparation of calcified tis-
sue for light microscopy," in Methods of Calcified Tissue Preparation, G. R.
Dickson (ed.), Elsevier Science Publishers B.V., Amsterdam, New York
Oxford, 1984, pp. 1-56.
30. R. K. Schenk and A. J. Olah, "Histomorphometrie," Handbuch der in-
neren Medizin VIIIA: Knochen, Gelenke, Muskeln, s. Kuhlencordt and
H. Bartelheimer (eds.), Springer Verlag, Berlin, Heidelberg, 1980,
pp. 437-494.
902 BUSER ET AL.

31. R. K. Schenk, “Cytodynamics and histodynamics of primary bone re-

pair,” in Fracture Healing, J. M. Lane (ed.), Churchill-Livingstone, New
York, 1987, pp. 23-32.
32. S. Steinemann, personal communication (1990).

Received October 31, 1990

Accepted February 19, 1991