France Lambert
Angelique Le
´
onard
Pierre Drion
Sophie Sourice
Paul Pilet
Eric Rompen
The effect of collagenated space filling
materials in sinus bone augmentation:
a study in rabbits
Authors’ affiliations:
France Lambert, Eric Rompen, Department of
Periodontology and Oral Surgery, Faculty of
Medicine, University of Liege, Lie
`
ge, Belgium
Angelique Le
´
onard, Department of Applied
Chemistry, University of Lie
`
ge, Lie
`
ge, Belgium
Pierre Drion, Central Animals Facility, Giga-R,
University of Liege, Lie
`
ge, Belgium
Sophie Sourice, Paul Pilet, INSERM, U 791,
Laboratory for Osteo-articular and Dental Tissue
Engineering, University of Nantes, Nantes, France
Paul Pilet, Service d’Odontologie conservatrice
Endodontie, CHU Nantes, Nantes, France
Corresponding author:
France Lambert
Service de Me
´
decine Dentaire
Domaine du Sart Tilman Bat B35
B-4000 Lie
`
ge
Belgium
Tel.: +32 473 655 755
Fax: +32 43 668 290
e-mail: france.lambert@chu.ulg.ac.be
Key words: bone regeneration, collagen, histomorphometry, hydroxyapatite, osteoconduction
Abstract
Aim: The inclusion of biomaterial particles used for alveolar bone regeneration in a carrier or in
binding agents such as collagen gel or fibers is of interest as a means to help with surgical
handling. However, the possible influence of collagen on bone tissue response to biomaterials is
poorly studied. The objective of the present study was to investigate, in a sub-sinus bone
augmentation model in rabbits, the effect of collagen at different stages of the osteogenesis
process. Histologic, histomorphometric and volumetric analyses were performed.
Materials and methods: Rabbits underwent a double sinus lift procedure using bovine
hydroxyapatite (BHA), collagenated bovine hydroxyapatite (BHAColl), and prehydrated and
collagenated porcine hydroxyapatite (PHAColl). Animals were sacrificed at 1 week, 5 weeks or
6 months. Samples were subjected to X-ray micro-tomography and histology. Qualitative analysis
was performed on the non-decalcified sections and quantitative histomorphometric analyses were
conducted using scanning electron microscopy (SEM). Volume variations of bone augmentations
were calculated at different time points.
Results: The three biomaterials allowed an optimal bone formation and were able to equally
withstand sinusal reexpansion. A comparable percentage of new bone, as well as 3D volume
stability, was found between the groups at each time po int. However, the PHAColl resorption rate
was significantly higher than the rates in other groups (P = 0.0003 ), with only 3.6% of the particles
remaining at 6 months. At 1 week, both collagenated groups displayed the presence of
inflammatory cells although BHA did not show any sign of inflammation. At 5 weeks and
6 months, the inflammatory process had disappeared completely in the BHAColl groups, whereas
some inflammatory-like cells could still be observed around the remaining particles of PHAColl.
Conclusions and clinical implications: Within the limitations of this study in rabbits, the findings
showed the presence of inflammatory-like cells at the early stage of bone regeneration when
collagenated xenogenic biomaterials were used compared to xenogenic granules alone.
Nevertheless, similar bone formation occurred and comparable 3D volumes were found at
6 months in the different groups.
Bone augmentation or preservation surgical
techniques are often used to preserve or
recreate an adequate bone volume for dental
implant placement (Esposito et al. 2006; Pje-
tursson et al. 2008). Autogenous bone graft-
ing was considered the gold standard for such
procedures because of its osteoinductive
properties. Nevertheless, autogenous bone
grafting has several disadvantages, such as
the need for a second surgical step and a vari-
able and unpredictable rate of resorption,
which led practitioners to consider alterna-
tive biomaterials (Sbordone et al. 2009). The
use of biomaterials as an osteoconductive
scaffold for bone formation in extraction
socket preservation, implant site develop-
ment (guided bone regeneration) or sinus lift
procedures are well documented today and
reliably used for several indications (Barone
et al. 2008; Chiapasco & Zaniboni 2009).
Most of the biomaterials used in alveolar
bone regeneration are available in particle
form and can be difficult to apply to the sur-
gical site. Some companies have developed
the inclusion of xenogenic particles in a bind-
ing collagenated agent to facilitate handling;
some have even made them injectable. Nev-
ertheless, the possible influence of collagen
Date:
Accepted 18 December 2011
To cite this article:
Lambert F, Le
´
onard A, Drion P, Sourice S, Pilet P, Rompen E.
The effect of collagenated space filling materials in sinus bone
augmentation: a study in rabbits.
Clin. Oral Impl. Res. 00, 2012, 1–7
doi: 10.1111/j.1600-0501.2011.02412.x
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
© 2012 John Wiley & Sons A/S 1
CLR
2412
B
Dispatch: 3.1.12 Journal: CLR
CE: Bharani
Journal Name Manuscript No.
Author Received: No. of pages: 7 PE: Karthik
on the bone tissue response to the biomate-
rial is poorly investigated in the literature
and remains controversial. Busenlechner
et al. showed a similar osteoconduction of
bovine hydroxyapatite (both mixed and
unmixed) with a carboxymethylcellulose and
collagen carrier after 6 and 12 weeks (Busen-
lechner et al. 2009). Nannmark et al. showed
that the addition of a collagen gel to colla-
genated porcine hydroxyapatite did not influ-
ence the bone tissue response to the material
after 2, 4 and 8 weeks post-insertion (Nann-
mark & Sennerby 2008). However, Araujo
et al. demonstrated that Bio-Oss
®
Collagen
(Geistlich Pharma AG, Wolhusen, Switzer-
land) obviously delayed the extraction socket
wound healing compared to regular socket
healing with a simple blood clot and showed
that inflammatory cells were present 3 days
and 1 week post-insertion (Araujo et al. 2009,
2010). Nevertheless, in those studies, non-
collagenated BHA was not investigated.
Comparative studies investigating the effects
of collagen at early stages of bone healing as
well as over the long term are therefore
needed.
The objective of this study was to qualita-
tively and quantitatively assess the early
bone formation process and mature bone
architectures of two different collagenated
xenogenic hydroxyapatite compared to bovine
hydroxyapatite alone, in a sinus lift model in
rabbits. Cell colonization, bone density,
osteoconductivity, resorption rate as well as
3D volume stability of bone augmentation
were explored.
Material and methods
Animals
New Zealand White rabbits (adult, males,
average body weight of 3 kg) were used in
the study. All experimental procedures and
protocols used in this investigation were
reviewed and approved by the Institutional
Animal Care and Use Ethics Committee of
the University of Lie
`
ge, Belgium. The “Guide
for the Care and Use of Laboratory Animals”,
prepared by the Institute of Laboratory Ani-
mal Resources, National Research Council,
and published by the National Academy
Press, was followed carefully.
Study design
This study is part of an overall project
where 96 sinus-lift procedures performed on
48 rabbits using 10 different types of space
fillers were assessed at three distinct time
points, 1 week, 5 weeks and 6 months,
respectively. Specifically, the space fillers
were allocated to the sinuses by a stratified
randomization and 16 rabbits were sacrificed
at each time point, so that at least three
sinuses were available for each space filler
at each time point, yielding a two-factor
experimental design (space filler and time)
with repeated measurements. The present
study focussed on the comparison of three
space fillers: a deproteinized Bovine
hydroxyapatite (Geistlich Bio-Oss
®
, Geistlich
Pharma AG) (BHA), a deproteinized bovine
hydroxyapatite incorporated in collagen
fibers of porcin origin (Geistlich Bio-Oss
®
Collagen, Geistlich Pharma AG) (BHAColl),
and a porcine hydroxyapatite still containing
the original collagen matrix and incorpo-
rated a 10% collagen gel (MP3
®
, Technoss,
Italy) (PHAColl). A total of 27 sinus-lift
procedures were analyzed from 26 different
rabbits.
Surgical procedure
Anesthesia of the rabbits was induced by
administration of a ketamine/xylazine bolus
(respectively 65/4 mg/kg, IM), 20 min after a
fentanyl/dehydrobenzperidol premedication
(0.22 ml/kg of a bolus 25 lg/1.25 mg/ml IM)
and 2 h before surgery, animals also received
buprenorphin at a dose of 0.05 mg/kg. This
was administered twice a day for 2 days.
Surgical interventions were performed under
strict sterile conditions. The surgical area
was shaved and disinfected with iodine, and
a straight incision was made to expose the
nasal bone and the naso-incisal suture lines.
The soft tissues were reflected with the
periosteum to access to the upper bone wall
of the sinus. Two ovoid windows (approxi-
mately 6 9 4 mm) were created bilaterally
using a round diamond bur. The membrane
was carefully raised from the floor and lat-
eral walls and the space-filling material was
inserted into the created compartment
(Fig. 1). The volume of filling material was
standardized to 0.4 ml per sinus. The bony
windows were covered with a resorbable
membrane (Biogide, Geistlich Pharma AG)
and the wounds were sutured with 4/0 poly-
ester thread (Permasharp, Hu Friedy, Rotter-
dam, The Netherlands). Animals were
sacrificed by injection of pentobarbital
(200 mg/kg, IV, after the same premedica-
tion as for surgeries). Samples were dissected
and soaked in fixative (6% formol). The sur-
gical procedures were performed by a single
operator.
Histological analysis
The samples were processed for non-decalci-
fied histology using polymethacrylate
(PMMA) resin. After fixation for about
1 week, the samples were dehydrated in
ascending graded ethanol series (24 h each
grade) and then placed in pure acetone for
24 h. Finally, samples were impregnated
with methylmethacrylate for 4 days with
one refreshment before embedding in
PMMA at 4°C for 4 days. Each resulting
non-decalcified block was cut sagittally with
a circular diamond saw (Leica, SP1600, Ger-
many
2
) at two different levels in the central
region. The first cut was in the area of the
window and the second 1.5 mm outward.
The two slices were then polished using a
grinding machine (Metaserv
®
2000, Buehler
3
)
and sputter coated with a thin layer of gold/
palladium on both sides. Samples were
observed under SEM (Leo 1450 VP
4
). SEM
observations were made using back-scattered
electron mode (BSE). Moreover, 30-lm sec-
tions were cut and polished using the same
material from the rest of each block in the
close vicinity of the central area and were
stained with HTX-eosin and counter-stained
with toluidine blue. To allow a better obser-
vation of cells, thin 7-lm sections were also
created using a hard tissue microtome (Leica
Fig. 1. Insertion of a Bio-Oss
®
Collagen block.
COLOR
2|Clin. Oral Impl. Res. 0, 2012 / 1–7 © 2012 John Wiley & Sons A/S
Lambert et al ! The effect of collagenated space filling materials
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
12h
10
Polycut SM 2500, Germany
5
) and stained
with Goldner trichrome.
Histomorphometry
Scanning electron micrographs (SEM, Leo VP
1450) were taken using the back-scattered
electron (BSE) mode at 309 magnification
and assembled to visualize the entire sinus.
These contiguous BSE pictures allowed a
quantitative evaluation of the mineralized
bone, the remaining biomaterial, and the soft
tissue areas based on their respective gray
levels using a semi-automatic image analyzer
(Leica Qwin, Germany
6
). The regions of inter-
est were manually defined, and the different
areas were automatically calculated. The fol-
lowing measurements were made: bone for-
mation, space filler area, and non-calcified
tissues, all expressed as percentages of the
augmented area;
X-ray microtomography analysis
All collected samples were first submitted to
x-ray microtomography. Before scanning, the
samples were transferred to an Eppendorf
®
7
tube containing fixative. The tube was
affixed to the brass stub and examined using
a Skyscan 1172 high-resolution desk-top
micro-CT system (Skyscan
®
, Kontich,
Belgium). The cone-beam source operated at
100 kV and 100 lA. The detector was a 2D,
1048 9 2000 pixel, 16-bit X-ray camera
8
. The
sample was rotated through 180° with a rota-
tion step of 0.49°, giving an acquisition time
of 30 min per sample. Taking into account
the camera definition and the source-object-
camera distance, 2D images with a pixel
size of 17.28 lm were obtained, using a
cone-beam reconstruction algorithm. The
corresponding 3D images were produced by
stacking all the 2D cross sections.
Analysis of the 3D images allowed the cal-
culation of the total volume of the regener-
ated space at baseline, at 5 weeks and at
6 months. The 3D measurements were car-
ried out using the CTscan software (release
2.5, Skyscan
®
, Kontich, Belgium).
Statistical analyses
Among the 26 rabbits, 23 (88%) had one
sinus included in the experimental design
and only three had their two sinuses
included, yielding a total of 29 sinus-lift
procedures. Hence, there was little loss in
(a)
(b)
(c)
(a)
(b)
(c)
(a)
(b)
(c)
Fig. 2. Histologic data observed with light microscopy for each studied space filler at 1 week. (a) Early steps of osteogenesis in a region close to the bone wall,29. (b) higher
magnification: note the presence of inflammatory cells in the BHAColl and PHAColl groups. (c) high magnification, 409: inflammatory cells were observed in the BHAColl
and PHAColl groups while mostly mesenchymal cells and fibroblasts were seen in the BHA group. (d) Images in the center of the sub-sinus created space, 209: collagen struc-
tures are distinguished in light green. (7 lm non-decalcified section, Goldner Trichrome staining).
COLOR
© 2012 John Wiley & Sons A/S 3|Clin. Oral Impl. Res. 0, 2012 / 1–7
Lambert et al ! The effect of collagenated space filling materials
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
efficiency by considering the 29 sinus as
independent statistical units. Results were
expressed as mean, standard deviation (SD),
minimum and maximum. The experimental
data were analyzed by two-way analysis of
variance (ANOVA) with repeated measure-
ments, allowing a test for interaction
between the two factors (time and space fil-
ler). When the interaction term was signifi-
cant, space fillers were subsequently
compared at each time point by one-way
ANOVA. Otherwise, the overall time and
space filler effects were tested. Results were
considered to be significant at the 5% level
(P < 0.05). A Bonferroni correction was
applied to account for multiple comparisons.
Statistical analyses were done using SAS ver-
sion 9.1 (SAS Institute, Cary, NC, USA).
Results
Descriptive analysis
One week after implantation (Fig. 2)
In the BHA group, no evidence of inflamma-
tion was found, whereas elongated mesen-
chymal and fibroblastic cells were found
everywhere in the cavity. Remnants of the
clot were almost invisible. The penetration
of connective and vascularized buds into the
sub-sinus space was observed along the bone
walls.
In the BHAColl group, a rich-cell tissue
(inflammatory-like) was observed in the
periphery of the cavity, along the bone walls
and under the lifted sinus membrane. The
center of the cavity was still poorly invaded
by the cells, whereas remnants of red blood
cells were seen among BHA particles and col-
lagen fibers. In the PHAColl group, rich-cell
tissue (inflammatory-like) were present in
the inter-particle areas all throughout the
cavity. Osteoclasts were observed along the
PHA particles.
Five weeks after implantation (Fig. 3)
In the BHA group, newly formed bone
bridged the particles of hydroxyapatite
together. Most of the particles’ surfaces were
in tight contact with a layer of new bone.
Only the center of the regenerated area was
not filled with any new bone. Some osteo-
clastic cells could be found along the parti-
cles, and osteoblastic cells were observed
only in the central region.
(a)
(b)
(c)
(a)
(b)
(c)
(a)
(b)
(c)
Fig. 3. Histologic data observed with light microscopy for each studied space filler at 5 weeks. (a) Newly formed bone was found in the periphery of the created space in the
three groups. However, in the BHA groups, bone colonization to the center areas seemed more advanced, 29. (b) higher magnification: Osteoclasts were seen in the three groups
but more predominantly in the PHAColl group. Note the presence of inflammatory cells localized in around the PHAColl granules. (c) Images in the center of the sub-sinusal
created space. (7 lm non-decalcified section, Goldner Trichrome staining).
COLOR
4|Clin. Oral Impl. Res. 0, 2012 / 1–7 © 2012 John Wiley & Sons A/S
Lambert et al ! The effect of collagenated space filling materials
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In the BHAColl group, newly formed bone
combined with a large number of capillaries
was observed along the bone walls, which
was nevertheless less extended to the center
than with BHA. Signs of inflammation and
collagen fibers were no longer observed.
In the PHAColl group, newly formed bone
also was observed along the bone walls. The
presence of small round cells as well as a
substantial amount of osteoclasts, was still
observed around the PHA.
Six months after implantation (Figs 4 and 5)
In the BHA group, bone marrow and adipo-
cytes were observed much more frequently
than at 5 weeks, while multinucleated cells
were not visible. Lamellar bone was found
solely in intimate contact with the particles
and was seen to bridge them together. Osteo-
blastic activity was very low, with neither
osteoid tissue nor osteoblasts being visible.
Bone trabeculae were covered with a unicel-
lular flat layer of cells.
In the BHAColl group, the anterior part of
the section mostly surrounded by the preexist-
ing bony walls displayed a similar architecture
as the BHA samples. However, the posterior
part, under the sinusale membrane, displayed
a less mature tissue, with a dense connective
tissue surrounding the bone particle network.
Samples treated with PHAColl also dis-
played a gradient of maturity from the
anterior to the posterior part of the sample,
similar to samples treated with BHAColl. In
the mature areas, PHA particles were com-
pletely resorbed and were replaced by rarefied
lamellar bone trabeculae displaying remodel-
ing activity; the marrow spaces were occu-
pied by adipocytes. In the less mature areas,
remnants of PHA granules surrounded by
small rounds cells and multinucleated osteo-
clasts were still observed and the non-calci-
fied tissue was of a dense fibrous tissue type.
MicroCT analysis: 3D volume variation
After 1 week, the mean volume of the aug-
mented tissue reached 344, 327, and
377 mm
3
for BHA, BHAColl, and PHAColl,
respectively. These were considered as the
baseline values. Two-way ANOVA applied to
the 3D volumes did not reveal any significant
interaction (P = 0.47) between space filler
and time. An overall negative time trend of
3D volume variation was observed
(P = 0.0041) mainly between 1 week and
5 weeks, although there was no effect of the
space filler (P = 0.34) (Table 1, Fig. 6).
Histomorphometric analysis
By applying a two-way ANOVA to the histo-
morphometrical data (bone formation, space
filler area and non-calcified tissues), bone for-
mation increased significantly with time
(P < 0.0001), although no difference was seen
among the three biomaterials (P = 0.84)
(Table 2). For the space filler area, a significant
interaction was observed between space filler
and time (P < 0.0001) (Table 2). Although this
parameter remained fairly stable for BHA and
BHAColl, a marked drop was observed for
PHAColl. After 6 months, the results were
significantly lower for PHAColl than those for
BHA and BHAColl (P = 0.0003). When consid-
ering the percentage of soft tissue, a signifi-
cant interaction was found between space
filler and time (P = 0.0002). Specifically,
although values between space fillers at
1 week were comparable (P = 0.23), they sig-
nificantly differed at 5 weeks (P = 0.0039) and
at 6 months (P = 0.0006). A decrease was
observed for BHA and BHAColl as opposed to
an increase for PHAColl.
Figure 6 shows the correlation between the
volume change over time as well as the per-
centage of new bone, percentage of space fil-
ler, and percentage of non-calcified tissues.
Discussion
The goal of the present study was to assess the
effect of collagenated xenogenic space fillers
on bone regeneration compared to the use of
(a)
(b)
(c)
(a)
(b)
(c)
Fig. 4. Histologic data observed with light microscopy for BHAColl and PHAColl at 6 months. (a) Junction between
the mature and immature bone, 7 lm non-decalcified section, 29. (b) higher magnification, 109: Mature areas dis-
played bone marrow with adipocytes as well as lamellar and rarefied mature bone. In the PHAColl groups, the
remodeling activity of the mature trabecula is well perceived with an osteoid zone covered by osteoblasts on one
side and osteoclasts on the other side. (c) Immature areas, 109: the non-decalcified space was filled with dense
fibrous tissue and large numbers of blood vessels. Some inflammatory cells are still punctually seen in the areas
where particles remained unresorbed.
COLOR
© 2012 John Wiley & Sons A/S 5|Clin. Oral Impl. Res. 0, 2012 / 1–7
Lambert et al ! The effect of collagenated space filling materials
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60