Home

Search

Collections

Journals

About

Contact us

My IOPscience

Evaluation of the tissue reaction to a new bilayered collagen matrix in vivo and its translation
to the clinic

This article has been downloaded from IOPscience. Please scroll down to see the full text article.
2011 Biomed. Mater. 6 015010
(http://iopscience.iop.org/1748-605X/6/1/015010)
View the table of contents for this issue, or go to the journal homepage for more

Download details:
IP Address: 134.93.123.250
The article was downloaded on 27/01/2011 at 12:40

Please note that terms and conditions apply.

IOP PUBLISHING

BIOMEDICAL MATERIALS

doi:10.1088/1748-6041/6/1/015010

Biomed. Mater. 6 (2011) 015010 (12pp)

Evaluation of the tissue reaction to a new

bilayered collagen matrix in vivo and its
translation to the clinic
Shahram Ghanaati1,2,8 , Markus Schlee3 , Matthew J Webber4 ,
Ines Willershausen5 , Mike Barbeck1 , Ela Balic6 , Christoph Gorlach6 ,
Samuel I Stupp7 , Robert A Sader2 and C James Kirkpatrick1
1

REPAIR-Lab, Institute of Pathology, Johannes Gutenberg University, Mainz, Germany

Clinic for Maxillofacial and Plastic Surgery, Johann Wolfgang Goethe University, Frankfurt Am Main,
Germany
3
Bayreuther Strasse 39, D-91301, Forchheim, Germany
4
Department of Biomedical Engineering, Northwestern University, Evanston, IL 60208, USA
5
Institute for Dental Material Sciences and Technology, University Medical Center of the Johannes
Gutenberg University, Mainz, Germany
6
Geistlich Pharma AG, Wolhusen, Switzerland
7
Department of Materials Science and Engineering, Chemistry, and Medicine, Northwestern University,
Evanston, IL 60208, USA
2

E-mail: ghanaati@uni-mainz.de

Received 24 August 2010

Accepted for publication 9 December 2010
Published 17 January 2011
Online at stacks.iop.org/BMM/6/015010
Abstract
This study evaluates a new collagen matrix that is designed with a bilayered structure in order
to promote guided tissue regeneration and integration within the host tissue. This material
induced a mild tissue reaction when assessed in a murine model and was well integrated within
the host tissue, persisting in the implantation bed throughout the in vivo study. A more porous
layer was rapidly infiltrated by host mesenchymal cells, while a layer designed to be a barrier
allowed cell attachment and host tissue integration, but at the same time remained
impermeable to invading cells for the first 30 days of the study. The tissue reaction was
favorable, and unlike a typical foreign body response, did not include the presence of
multinucleated giant cells, lymphocytes, or granulation tissue. In the context of translation, we
show preliminary results from the clinical use of this biomaterial applied to soft tissue
regeneration in the treatment of gingival tissue recession and exposed roots of human teeth.
Such a condition would greatly benefit from guided tissue regeneration strategies. Our
findings demonstrate that this material successfully promoted the ingrowth of gingival tissue
and reversed gingival tissue recession. Of particular importance is the fact that the histological
evidence from these human studies corroborates our findings in the murine model, with the
barrier layer preventing unspecific tissue ingrowth, as the scaffold becomes infiltrated by
mesenchymal cells from adjacent tissue into the porous layer. Also in the clinical situation no
multinucleated giant cells, no granulation tissue and no evidence of a marked inflammatory
response were observed. In conclusion, this bilayered matrix elicits a favorable tissue reaction,
demonstrates potential as a barrier for preferential tissue ingrowth, and achieves a desirable
therapeutic result when applied in humans for soft tissue regeneration.
(Some figures in this article are in colour only in the electronic version)

Author to whom any correspondence should be addressed.

1748-6041/11/015010+12$33.00

2011 IOP Publishing Ltd Printed in the UK

Biomed. Mater. 6 (2011) 015010

S Ghanaati et al

membrane stability, several cross-linking techniques have

been applied [11, 1619]. While this leads to increased
mechanical stability of the membrane [11, 16, 19], it also
inhibits the attachment and proliferation of human periodontal
ligament fibroblasts and osteoblasts compared to native
collagen [20]. Cross-linked collagen has also been reported
to induce a severe foreign body reaction in vivo [21]. As a
consequence, alternative processing techniques for collagen
have been developed. One such technique involves the
combination of both non-cross-linked native collagen III,
which undergoes a relatively fast degradation, and collagen I,
which is more resistant, in order to regulate the biodegradation
of membranes applied in GTR. Recently, a bilayered nonR
cross-linked collagen-I/III membrane (BioGide
) implanted
subcutaneously in rats was demonstrated to undergo a
homogeneous and transmembranous vascularization after 2
weeks and persisted in the tissue for up to 4 weeks [21,
22]. It was also shown that membrane thickness continuously
decreased, with a breakdown 4 weeks after implantation. A
fast transmembranous vascularization, while enhancing the
tissue integration of the scaffold, at the same time results in a
loss of the barrier function of the material. This could prove
detrimental to applications in GTR.
The aim of this study was to assess the tissue reaction to
R
Mucograft
(MG), a new non-cross-linked collagen matrix. It
is composed of collagen type I and type III without further
cross-linking or chemical treatment. The MG matrix is
bilayered with one side being thin, smooth and of low-porosity,
while the other is a more porous three-dimensional network
(figure 1). In this histological study, special emphasis was
placed on the comparative tissue integration on both sides of
the matrix, focusing on host tissue reaction to the material
at each of its different surfaces and the cells involved in this
reaction. Additionally, the material thickness was measured
histomorphometrically throughout the time points of the study.
Finally, we compare our observations developed in a murine
model to those from a pilot human application of MG to treat
soft tissue defects within the oral cavity, so as to evaluate how
the information gained from our small animal study translated
directly into the observations from the use of MG as a matrix to
therapeutically reconstruct human gingival soft tissue. Thus,
it was hoped to test the validity of the animal model for clinical
translation purposes.

1. Introduction
The loss of soft tissue in the oral cavity caused by either
receding gingival tissue or scar tissue, in addition to being
unaesthetic, can be a tremendous source of pain and suffering.
Current gold-standard tissue transfer procedures involve
transplanting an oral autograft, but these can cause additional
discomfort and are accompanied by a risk of fatality when
tissue is harvested from other parts of the oral cavity such
as the palate due to the proximity of major blood vessels.
Synthetic strategies, including the application of biomaterial
membranes, have been developed to combat this within
oral and maxillofacial surgery over the past two decades
using concepts from the so-called guided tissue regeneration
(GTR). In GTR, the goal is to promote specifically healthy
connective tissue ingrowth, while inhibiting that of fibrotic
scar tissue. This is akin to guided bone regeneration (GBR),
where a defect in bone is treated by similar means in
order to facilitate bone ingrowth within this defect, while
minimizing ingrowth of reactive fibrotic tissue. Membranous
biomaterials have been applied to promote preferentially the
ingrowth of cells involved in the tissue regeneration process,
while preventing ingrowth from undesired tissue into the
defect [15]. For a successful outcome, barrier membranes
used in GTR must possess certain properties, including
biocompatibility, preferential tissue integration, place-holder
characteristics, and physiochemical stability [2, 6]. The
first generation of such membranes were primarily made of
poly(tetraflouroethylene) (ePTFE) [7], a polymer with high
stability in biological systems. However, this non-resorbable
membrane requires a second intervention for its retrieval,
accompanied by further risk of infection, potential damage
to the newly regenerated tissue, and an increased likelihood
of further intervention [8, 9]. Thus, the drawbacks associated
with retrieval have led to the development of bio-resorbable
barrier materials. Bio-resorbable polyesters with different
biodegradation properties have been developed for various
applications in reconstructive surgery including GTR [10
12]. However, no satisfying outcome was achieved with these
membranes, as they exhibited a fast degradation related to a
foreign body reaction, which did not permit the desired control
of tissue regeneration.
Collagen-based materials have also been explored for
applications to GTR strategies.
Collagen is the most
abundant family of proteins in the human body and
is physiologically ubiquitous.
Moreover, neutrophils,
monocytes, and fibroblasts recruited during wound healing
release matrix metalloproteases (MMP), which results in
enzymatic biodegradation of collagen [13]. Its natural
origin combined with its relative ease of biodegradation
make collagen a broad candidate for biomaterial applications.
Collagen type I is also known to have angiogenic potential
[6, 14, 15], a characteristic that makes it further desirable to
promote the ingrowth of healthy tissue. In applications as
membranes, however, the biodegradation of collagen proved
to be a major disadvantage, since it limited the time scale
over which the membrane exhibited a barrier function. In
order to decrease the rate of collagen degradation and enhance

2. Materials and methods

R
2.1. Production of Mucograft
R
Mucograft
(MG) is a pure collagen matrix obtained by a
proprietary standardized manufacturing process and sterilized
by gamma irradiation. The device has been CE-marked
and 510(k) FDA-approved. The matrix is made of collagen
type I and type III without further cross-linking or chemical
treatment. The collagen is extracted from a veterinary certified
porcine source and is carefully purified to avoid antigenic
reactions. The collagen is processed into a bilayered matrix
with one side being a thin, smooth and low-porosity compact
layer (CL) and the other a more porous three-dimensional

Biomed. Mater. 6 (2011) 015010

S Ghanaati et al

(A)

(B)

(D)

(C)

(E)

Figure 1. Scanning electron microscopy of the MG scaffold, showing a low magnification cross section (A). In addition, the face (B, left)
and cross-section (C) of the low-porosity compact layer of this scaffold is shown, along with the face (D) and cross-section of the more
porous spongy layer (E). CL = compact layer and SL = spongy layer.

spongy layer (SL) (figure 1). The low-porosity surface, made

from porcine peritoneum, has elastic properties that permit
suturing to the host mucosal margins. The porous surface,
derived from porcine skin, allows tissue adherence favoring
wound healing and promoting cell integration. The porosity
is obtained through defined parameters and a controlled
lyophilization process. This side is turned toward the bone
defect and/or soft tissue to encourage bone-forming cells and
tissue growth and to stabilize the blood clot. Both layers are
combined through a biophysical interweaving process without
any chemical manipulation. The volume fraction of pores in
the matrix is 90% and the size distribution for these pores
ranges from 5 to 200 m, with smaller pores being primarily
located on the compact layer and larger pores found in the
spongy layer.

2.2. Scanning electron microscopy

Scanning electron microscopy (SEM) was performed at
the Northwestern University Electron Probe Instrumentation
Center (EPIC, Evanston, IL, USA), using a Hitachi S4800
scanning electron microscope with a 5 kV accelerating voltage.
To prepare samples for imaging, MG was first hydrated in PBS
and subsequently dehydrated in ethanol. It was then dried at
the critical point, cut to reveal a cross section, and coated with
roughly 9 nm OsO4 prior to imaging.
2.3. Murine experimental design
The Committee on the Use of Live Animals in Teaching
and Research of the State of Rhineland-Palatinate, Germany,
approved these studies. In total, 30 female 5-week-old
CD-1 mice were obtained from Charles River Laboratories,
3

Biomed. Mater. 6 (2011) 015010

S Ghanaati et al
R
2.7. Pilot human Mucograft
use

Germany. The animals were maintained for 1 week before

use at the Laboratory Animal Unit, Institute of Pathology,
Johannes Gutenberg University of Mainz, Germany, as
follows. They were fed with regular mouse pellets (Laboratory
Rodent Chow, Altromin, Germany), water was available ad
libitum and there was an artificial lightdark cycle of 12 h
each. The mice were randomly distributed into two groups.
The first group (n = 4 animals/time point) was implanted with
MG while the second group (n = 2 animals per group) served
as a sham-operated control with no biomaterial implantation.
The time points evaluated were 3, 10 15, 30 and 60 days.
Sterile MG samples of 10 10 cm were implanted with the
spongy side facing down into preformed subcutaneous pockets
of animals rostral subscapular region, according to a previously
described method [23]. All animals survived the operation and
subsequent evaluation period without complications.

A clinical application of MG was performed at the private

practice of one of the co-authors (MS). In this paper, we
provide evidence from this trial only in so far as to confirm our
findings in vivo and thus support a possible clinical translation
R
of Mucograft
. A future report, detailing the full clinical
measures and outcomes of this study, will follow. In this
study, a total of 11 patients (7 females, 4 males) who had
at least one buccal recession classified as Miller class I or
II were selected and each patient gave written consent for
outcomes of this study to be used anonymously for research
and publication purposes in accordance with local ethical
guidelines. In total, 54 recessions were treated with MG.
Patients were otherwise healthy and were verified to have
no allergy to porcine products and no known anaphylactic
reaction. After local anesthesia, the exposed root was scaled
and planed to the bottom of the pocket with rotating burs,
ultrasonic instruments and curettes. Deeper instrumentation
was avoided to prevent fibrous attachment. The root was
flattened, smoothened and decontaminated. No chemical
root conditioning was performed. After sulcular incision,
a coronal displaced split thickness flap without releasing
incision was performed, according to the incision method
previously described [25]. The flap preparation, in contrast
to Zucchelli, was a complete split thickness flap. The flap
was considered to be immobilized enough when it stayed
passively at a level coronal to the cemento-enamel junction.
Beside complete root coverage, the aim of this surgery was
to enhance the thickness and stability of gingival tissue. The
MG collagen matrix was placed underneath the prepared flap.
After rehydration, the matrix becomes flexible, remains stable,
and is readily sutured. The matrix was adapted in size
and fixed by sling sutures around the recipient teeth. The
matrix was trimmed to reach at least a level 2 mm apical
the bony margin. The coronal position was 1 mm below
the cemento-enamel junction. The knots were placed on the
lingual aspect, making removal painless and atraumatic. The
coronal advanced flap completely covered the matrix and was
also fixed by sling sutures to accomplish a precise adaptation
around teeth. Following surgery, patients were instructed to
avoid mechanical trauma to the wound. Sutures were removed
2 weeks after surgery. Patients were recalled at time points
following this for a professional supragingival tooth cleaning
and remotivation.
One patient in this study was recalled at 6 weeks for
R
complications unrelated to the Mucograft
scaffold itself, but
to make adjustments to the surgical procedure. In the course
of this procedure, a biopsy was obtained which allowed a
histological evaluation of the performance of the MG scaffold
at this point. Again, written consent was received from the
patient to use the results of this biopsy for research publication
purposes. The results from this biopsy, and their connection
to the results we have seen in the subcutaneous implantation
model, will be the primary focus for this report. Tissue was
processed for histology by similar means to that described
for murine tissue and the tissue was stained using H&E and
Alizarin Red.

2.4. Tissue preparation for histology and

immunohistochemistry
After explantation at the specified time points, tissue was
processed for histological analysis as has been previously
described [23]. Sections from each region were stained with
Mayers hematoxylin and eosin (H&E), Azan, Sirius Red,
Ladewig, and Movats Pentachrome. Murine vessels and
macrophages were detected immunohistochemically by means
of an anti-rabbit polyclonal CD31 antibody (GeneTex, Inc.,
USA) and a monoclonal F4/80 antibody (eBioscience, San
Diego, USA), respectively. The DAKO REALTM EnVisionTM
detection system (DAKO, Glostrup, Denmark) was used for
visualization. Counterstaining of the immunohistological
samples was performed with hemalaun.
2.5. Morphological evaluation of the inflammatory response
Histopathological evaluation was carried out using a
previously described method [23, 24]. Briefly, cellular
inflammatory component (i.e. neutrophils, lymphocytes,
plasma cells, macrophages, giant cells) and biomaterial
vascularization and degradation were qualitatively and
quantitatively evaluated by light microscopy of histological
sections and software quantification tools.
2.6. Thickness measurements
Thickness measurements of the matrix were done using
previously described histomorphometric methods [23, 24].
Briefly, the region of interest containing the MG biomaterial
and the corresponding peri-implant tissue was used to generate
of a total scan which assembles 100120 images into one large
image at 100 magnification. For each animal the thickness
of both the compact layer and the spongy layer was measured
at each time point in order to assess the dynamic degradation
of these components over time. The material was measured
at up to 15 points along the histological cross-section for
each animal at each time point, and the total thickness was
calculated from the mean of these measurements. This mean
was taken as the total thickness for the layers of the material.
4

Biomed. Mater. 6 (2011) 015010

S Ghanaati et al

3. Results
(A)

3.1. SEM of the scaffold

SEM of a cross-section of the MG scaffold revealed the
morphology of the bilayered matrix (figure 1(A)). The
compact layer (CL), designed to be less cell permeable, was
found to have a very low-porosity surface (figure 1(B)), while
the cross-section of the matrix revealed that this layer consisted
of large parallel sheets of densely compressed collagen fibers
(figure 1(C)). The spongy layer (SL), designed to facilitate
tissue ingrowth, had a surface that was highly porous and
had more randomly aligned and diffusely packed collagen
fibers (figure 1(D)), with a cross-section of this region having
very similar structural characteristics (figure 1(E)). Overall, the
structural characteristics of the matrix suggest it has potential
for use as a GTR scaffold, with two very structurally different
layers that would, presumably, promote preferential tissue
infiltration. The cross-sectional thickness of the membrane
in SEM was of the order of 1.61.8 mm, with a compact layer
of roughly 0.4 mm and a spongy layer of roughly 1.3 mm.

(B)

3.2. Murine histological results

Animals implanted with MG showed no signs of hemorrhage
or necrosis at the implantation site. At 3 days following
implantation, there was no evidence of capsule formation
around the material. The cells in the peri-implant tissue on
either side of the matrix were mainly fibroblasts with a few
macrophages, while there was almost no observed neutrophil
infiltration (figure 2(A)). Histologically, both the compact
and spongy layers of the MG scaffold were easily detectable.
Both components were well integrated within the implantation
bed (figure 2(A)). At this early time point, tissue ingrowth
was already observed from the spongy surface of the matrix,
indicated by evidence of several infiltrating cells (figure 1(B)).
However, on the compact surface of the matrix, a cell-rich
connective tissue was observed adjacent to the matrix which
enabled cell attachment but appeared to simultaneously inhibit
the ingrowth of tissue into the matrix, as few cells could be
seen infiltrating this side (figure 1(C)).
At day 10, the two components of the matrix were still
distinctly visible by histology (figure 3(A)). The compact layer
still appeared to inhibit the penetration of the peri-implant
cells into the core of the matrix, but the beginnings of an
ingrowth of peri-implant cells were evident within the layers
of the dense component at this time (figure 2(B)). The tissue
ingrowth into the spongy layer that was originally observed at
day 3 had progressed, with more cells reaching deeper within
the scaffold and beginning to secrete their own matrix within
the collagen fibrils (figure 3(C)). Immunohistological staining
with F4/80 showed the presence of macrophages within the
matrix and suggests their possible role in the degradation of
the biomaterial (figure 3(D)).
At day 15, the tendency for tissue ingrowth that was seen
on day 10 for both sides of the matrix continued (figure 4(A)).
The two layers of the matrix also remained clearly visible
by histology. The compact layer still serves as a barrier for
cell ingrowth into the core of the matrix, with some cells

(C )

Figure 2. Post-implantation day 3 histological images showing a

cross-section of the scaffold with Azan stain at 100 (A) as well as
the interface between the host tissue and the spongy layer with H&E
stain at 400 (B) and the corresponding interface for the compact
layer with H&E stain at 400 (C). Cells (arrows) can be seen
penetrating the spongy layer, which the compact layer shows very
little cell infiltration. SL = spongy layer, CL = compact layer and
ST = subcutaneous tissue. All scale bars are 100 m.

being diffusely spread through the thick sheets of collagen

(figure 4(B)). The spongy layer was invaded by peri-implant
tissue, with a further increase in matrix production by these
host cells (figure 4(C)). Additionally, immunohistochemical
studies showed an increase in macrophages in this spongy
layer compared to day 10 (figure 4(D)).
By day 30 after implantation, significant changes
were evident in the histology of the matrix (figure 5(A)).
Specifically, at this time the compact layer of the matrix no
longer maintained its function as a non-penetrable barrier, as
cells could be seen to have traversed this layer and entered
the deeper regions of the scaffold. At this time, cells had
penetrated the scaffold from both surfaces of the matrix and
became more dispersed throughout (figures 5(B) and (C)).
5

Biomed. Mater. 6 (2011) 015010

S Ghanaati et al

(A)

(B)

(C)

(D)

Figure 3. Post-implantation day 10 histological images showing a cross-section of the scaffold with H&E stain at 100 (A) as well as the
interface between the host tissue and the compact layer with H&E stain at 400 (B) and the corresponding interface for the spongy layer
with Movats pentachrome stain at 400 (C). Cells continue to penetrate the spongy layer and secrete their matrix, while the compact layer
continues to inhibit cell penetration of the scaffold, though cells can be seen between the sheets of collagen in this layer. F4/80 staining for
macrophages in the spongy layer (D) reveals some at the surface and beginning to enter the scaffold at 400. SL = spongy layer, CL =
compact layer and ST = subcutaneous tissue. All scale bars are 100 m.

spongy layer on subcutaneous implantation, scaffold wetting,

or potential histological artifacts. The thickness of the compact
layer is similar regardless of technique. Over the course of
implantation, the thickness of the spongy layer decreased to
289 68 m, a decrease of 40% from the day 3 thickness.
The compact layer, on the other hand decreased considerably
in thickness to 137 33 m by day 60, a decrease of almost
73% compared to its value on day 3. The MG scaffold as a
whole had a thickness of 426 62 m at day 60, representing a
decrease in thickness of 57%. The largest change in thickness
was in the compact layer between days 10 and 15, which
preceded the breakdown of this layer as a barrier by day 30.

At day 60 after implantation, the dense component of

MG was much less visible and what remained of the spongy
layer of the matrix was invaded by a vascularized connective
tissue (figure 6(A)) and had become very well integrated within
the implantation bed (figures 6(B) and (C)). There were no
multinucleated giant cells and only a few lymphocytes were
observed within the implantation bed throughout the entire
observational period, with the degradation of the material
being primarily performed by macrophages. Accordingly, at
no point in the histological study did we observe a typical
foreign body response.
3.3. Histomorphometric assessment of thickness
Using histomorphometric measurements of a total scan of the
biomaterial, we were able to track the total MG thickness
as well as the thickness of each bilayer, as these layers
were distinctly visible through histology due to their very
different morphologies (figure 7). These measurements
revealed that the compact layer was more inclined to be
thinned by invasion from host cells than the spongy layer.
The total scaffold thickness at day 3 was 980 90 m,
consisting of roughly the same thickness of a spongy layer
(479 80 m) and a compact layer (501 46 m). Both
the total thickness and the thickness of the spongy layer
differed from those seen in SEM of the scaffold, possibly
as a result of factors arising from the compression of the

3.4. Scaffold vascularization

Using CD31 staining, we evaluated the material for its effect
on vascularization of the peri-implantary tissue, particularly
at the tissue interface of the spongy layer of the material. At
3 days following implantation, there was some evidence of
vessels in the peri-implant space (figure 8(A)), but this did not
appear to be a significant increase from normal physiological
levels. At 10 days, there were still only a few vessels found
in the peri-implant region (figure 8(B)). By 30 days following
implantation, a few vessels were observed within the scaffold
periphery (figure 8(C)) and this did not change considerably
by day 60 (figure 8(D)), with the center of the scaffold being,
6

Biomed. Mater. 6 (2011) 015010

S Ghanaati et al

(A)

(B)

(C)

(D)

Figure 4. Post-implantation day 15 histological images showing a cross-section of the scaffold with Ladewig stain at 100 (A) as well as
the interface between the host tissue and the compact layer with Azan stain at 400 (B) and the corresponding interface for the spongy layer
with Movats Pentachrome stain at 400 (C). Cells continue to penetrate the spongy layer and secrete their matrix, while the compact layer
continues to inhibit cell penetration of the scaffold, though cells can be seen between the sheets of collagen in this layer. F4/80 staining for
macrophages in the spongy layer (D) reveals even more at the surface and beginning to enter the spongy layer of the scaffold at 400. SL =
spongy layer, CL = compact layer and ST = subcutaneous tissue. All scale bars are 100 m.

procedure.
The histology of this biopsy corroborated
the findings in the murine evaluations.
We observed
a distinct difference in the tissue reaction of the two
layers (figure 9(D)), with the spongy layer demonstrating
good tissue integration with the surrounding connective
tissue and substantial cell infiltration, while the compact
layer appeared to be much less infiltrated with cells,
though it still remained integrated in the host tissue at
its surface. The spongy layer was also well integrated
within the neighboring connective tissue (figure 9(E)).
As was seen in the murine model, there was no evidence of
an adverse tissue reaction, with no multinucleated giant cells
and no lymphocytes present. Also, there was minimal scaffold
vascularization, consistent with our findings of a low level of
vascularization in the murine model. The results from this
biopsy closely mirrored those from our in vivo evaluation in
terms of tissue reaction and preferential tissue ingrowth.

at most, sparsely vascularized. Overall, at no point was there a

dramatic increase in vascularization in the peri-implant tissue
and the scaffold itself was minimally vascularized, beginning
at the periphery and progressing slowly to the center.
R
3.5. Human Mucograft
pilot study

The murine in vivo studies demonstrated a very favorable tissue

reaction as well as preferential control over tissue ingrowth,
with cells penetrating the scaffold from the spongy layer
before they were able to penetrate the compact layer. This
points to the possibility of MG scaffolds as a barrier for soft
tissue regeneration. As mentioned, one clinical need for such
materials is for the treatment of exposed gingival roots in
teeth. Thus, MG was evaluated for its ability to serve as
a barrier and promote tissue ingrowth in this application.
Shown here is an example from a single patient with an
exposed root from receding gingiva (figure 9(A)) and who
received MG as therapy. The gum was retracted to make a
flap and MG was inserted so that the spongy side was facing
the tooth (figure 9(B)). The flap was then sutured in place
over the MG scaffold. A follow-up after treatment revealed
that the gum level has returned to normal height, with the
color and appearance matching that of healthy gingival tissue
(figure 9(C)). These results suggest the potential for MG to be
successfully translated to the clinic in this application.
At 6 weeks following treatment, a biopsy of the scaffold
and surrounding tissue was collected as part of another

4. Discussion
In this in vivo evaluation, we examined the tissue reaction to
a bilayered porcine collagen I/III matrix with a compact and
spongy side. Special emphasis was placed on the integration
of this material within the implantation bed for each of its
two morphologically distinct surfaces. In the animal study,
no capsule formation was observed around the implanted
matrix and there was no acute accumulation of neutrophils
7

Biomed. Mater. 6 (2011) 015010

S Ghanaati et al

(A)

(B)

(C)

Figure 5. Post-implantation day 30 histological images showing a cross-section of the scaffold with Azan stain at 100 (A) as well as a
higher magnification image with H&E stain at 200 (B) and a total scan of a histological slide with H&E stain at 100 magnification
(C) showing cells that have progressed to the center of the scaffold from both sides, indicating a loss of barrier function for the compact
layer. SL = spongy layer, CL = compact layer and ST = subcutaneous tissue. All scale bars are 100 m.

and macrophages following 3 days of implantation. The

cellular degradation of the biomaterial was initiated primarily
by macrophages, with no evidence of multinucleated giant
cells on either matrix surface at any time point of the study.
Our findings suggest that there are no issues pertaining to the
porcine collagen source, as there was no sign of any adverse
immunological reaction within the host, nor was there evidence
of a foreign body reaction at any point in the study. It is known
that giant cell formation is dependent on biomaterial surface
morphology, requiring the adsorption of a specific spectrum
of proteins in order to trigger fusion of adherent mononuclear
cells into multinucleated giant cells [2527]. As no giant
cells were seen, the study demonstrates that macrophages
are sufficient for biodegradation of the MG matrix. Our
present findings are in agreement with observations for similar
materials in other in vivo studies that have demonstrated good
tissue integration and shown no foreign body reaction for noncross-linked porcine derived collagen I/III membranes [21].
In this study, the center of the MG matrix did not
become vascularized until some point between days 30 and
60, which contrasts with studies in similar collagen I/III
materials that demonstrate vascularization 2 weeks following
implantation [22]. This is most likely attributable to the
differences in material processing for the MG matrix compared
to this other collagen I/III membrane. We have previously
shown that changes in material processing for biomaterial
scaffolds can have a significant effect on the resulting tissue
reaction [23], so that differences in processing methodology
could certainly account for the variance between our results

and those published for other collagen-I/III materials. The

slow vascularization of MG is likely a result of the barrierlike function of the compact collagen layer, which did not
allow cell infiltration until 30 days following implantation.
Vascularization of the spongy layer began at 2 weeks, which
is in accordance with published reports. The degree to which
the material is vascularized is low, perhaps on account of the
relatively small amount of granulation tissue formed. We
have previously shown that similar subcutaneous implantation
of silk fibroin (SF) leads to an active granulation tissue
within the implantation bed and corresponds to a higher
degree of scaffold vascularization [23]. Although SF is
also a natural-based biomaterial, its implantation evoked the
activation of macrophages and their fusion into multinucleated
giant cells within the implantation bed. These cells, which
were found on the surface of the SF fibers, were involved in
degradation of the SF scaffold and their number decreased as
the material was degraded. Multinucleated giant cells are a
known participant in the foreign body response and arise from
macrophage fusion. Since SF and collagen are both naturally
derived protein-based materials, but yet have different levels
of vascularization, it is possible that multinucleated giant cells
play a role in scaffold vascularization as part of the foreign
body response. Our hypothesis is that the host tissue reaction
is, in some way, related to the ease with which the material is
degraded. Materials based on collagen, which is ubiquitous
in mammalian tissue, may not activate the degree of cellular
immune response that a naturally derived but foreign material
such as SF will, perhaps due to the structural similarity of
the implanted collagen to physiological tissue. However,
8

Biomed. Mater. 6 (2011) 015010

S Ghanaati et al

(A)

(B)

Figure 7. Histomorphometric analysis of the total thickness of the

scaffold as well as the thickness of each of the two layers of the
scaffold over time of implantation.

as a place holder and barrier for a longer period of time, an

ideal characteristic in GTR applications.
We also focused on the materialtissue interface. After
implantation the compact layer of the matrix served a barrier
function, preventing cellular infiltration from the surrounding
tissue. Although peri-implant cells were able to intercalate
these layers of collagen in this part of the matrix by day 10, they
were not able to penetrate this layer completely until 30 days
after implantation. These results emphasize that the compact
layer promotes cell adhesion and tissue integration, but that
invading cells need up to 30 days in order to traverse this layer.
On the spongy layer of the MG matrix, a slow but continuous
cell and tissue ingrowth was observed through the porosity of
this surface. This spongy side, while allowing cells to infiltrate
without much resistance, maintained its structure through the
entirety of the study and became integrated into the host tissue
as opposed to undergoing degradation. The porosity allowed it
to act as a scaffold on which peri-implantary cells, connective
tissue, and microvessels were able to grow and advance to the
center of the matrix.
Our results from the preliminary human application
of MG are exciting, as the matrix demonstrated success
in reversing the recession of gingival tissue, resulting in
more healthy tissue with a raised gum line in the patient.
The aesthetic results are quite remarkable. Furthermore,
histological evidence demonstrates that within human tissue,
the compact layer of the matrix again serves a barrier function.
This gave the periodontal defect the time needed to regenerate
by maintaining isolation from the fibrous tissue and oral
mucosa for at least 43 days. Along with this, the material
showed no adverse tissue reaction, an ideal characteristic of
a material to be applied clinically. These results highlight
the functional performance of the MG matrix in a truly
translational context and emphasize the suitability of this
material for application in GTR.
One impressive feature of the combined animal and the
preliminary clinical study is that the human histology from

(C)

Figure 6. Post-implantation day 60 histological images showing a

cross-section of the scaffold with sirius red stain at 100 (A) as
well as higher magnification images of the material-tissue interface
with Ladewig stain and H&E stain at 200, (B) and (C). The
material is totally invaded by cells and appears more as a connective
tissue, though there is only some evidence of blood vessels (arrow in
C). The spongy layer of the material is well integrated within this
new connective tissue. SL = spongy layer and ST = subcutaneous
tissue. Arrow heads in (A) show the materialtissue interface, while
the double-headed arrow demonstrates the spongious layer of the
matrix. All scale bars are 100 m.

even among collagen scaffolds, the tissue reaction may vary

depending on such parameters as scaffold thickness, porosity,
chemical cross-linking, source, and purification methods. In
this study very little inflammation was seen and, perhaps
correspondingly, a low level of vascularization. However,
the density of vascularization is more in accordance with
the physiological levels for this tissue site as was observed
for sham-operated animals [28], which in some applications
for biomaterials may be desirable when there is no need for
a vascularization in excess of the particular host tissue site.
Moreover, the low level of vascularization also likely results
in a slower material degradation, allowing the material to serve
9

Biomed. Mater. 6 (2011) 015010

S Ghanaati et al

(A)

(B)

(C)

(D)

Figure 8. Immunohistological staining for CD31 in order to visualize blood vessels (arrows) from scaffolds implanted for 3 days (A),
10 days (B), 30 days (C), and 60 days (D), all at 400. At most, a mild vascularization is evident in the peri-implantary tissue and within the
MG scaffold itself. All scale bars are 100 m.

(A)

(D)

(B)

(E )

(C)

Figure 9. Results from the clinical study applying the MG scaffold to treat gingival tissue recession (arrows), showing one such recession
(A) that has been treated by the implantation of the MG scaffold (B) with the compact layer facing out in the image. Following treatment,
the gingival tissue level has been restored to normal height (C). Histological results from a biopsy obtained in the course of the study
demonstrate similar findings as those in the in vivo murine model, with the compact layer serving as a barrier, while the spongy layer is
infiltrated with cells and becomes transformed into a connective tissue, demonstrated with Alzarin red stain (D) and H&E stain (E) at 400.
SL = spongy layer, CL = compact layer and CT = connective tissue. All scale bars are 100 m.

10

Biomed. Mater. 6 (2011) 015010

S Ghanaati et al

our pilot evaluation corroborated our results when the material

was tested in the subcutaneous murine model. It is one thing to
evaluate a material in vivo and speculate based on these results
how the material would perform in a therapeutic context. In
this study we have direct evidence that our predictions for
the performance of the material formulated in vivo translate
very well to observations in the clinical setting. The time
frame over which the barrier function of the compact layer
is maintained does differ, owing perhaps to the differences
in species or in tissue site. However, regardless of whether
MG was placed subcutaneously within a mouse or within
human gingival tissue, the matrix maintained its ability to
promote preferential tissue ingrowth. Moreover, both studies
reveal a favorable tissue reaction and a physiological level
of scaffold vascularization. It is evident that more extensive
clinical studies are required, and these are currently being
conducted.
The results from the pilot clinical evaluation support
the findings from our animal model, indicating that the
subcutaneous implantation model can appropriately predict
the tissue reaction and preferential tissue ingrowth for such
materials. This is an exciting finding, as it is especially
satisfying to know that the sacrifice of animals was not without
purpose and that the lessons learned in these animal models
translated to clinical findings. In addition, observations in the
animal model would suggest that the application of MG is not
limited to only the treatment of gingival recessions. Although
such an application has been shown here to exemplify the
clinical applicability of the material. It could certainly find use
in applications to promote the ingrowth of healthy tissue in the
other locations within the oral cavity, used as a treatment for the
re-epilthelialization of chronic or non-healing wounds, or used
in a number of other applications for soft tissue regeneration
where a preferential ingrowth and a barrier to the formation of
fibrotic tissue are essential.
The results of this study underline the importance
of material composition and morphology for use as GTR
scaffolds. The material evaluated here demonstrates ideal
functionality for this role in vivo. The porous side of the
MG matrix permits cells to integrate and grow into the center
region of the scaffold. On the other hand, the compact layer
inhibits connective tissue ingrowth. These features allow for
preferential cell ingrowth, a vital characteristic of scaffolds
designed for regeneration of soft tissue. Additionally, the
MG matrix demonstrates an excellent tissue reaction, resulting
in no multinucleated giant cell formation and becoming well
integrated into the peri-implant tissue. These studies give
promise for the translation of MG as a biomaterial, a promise
that is realized in the example from a pilot human application
using MG to treat receding gingival tissue. In this example, the
MG matrix performed its function both as a barrier material
and also as a structural component within which functional
tissue was constructed. The results from this study are
congruous with those of our murine studies, and support our
conclusions from these studies that MG is an excellent option
to promote the preferential regeneration of soft tissue.

5. Conclusion
In this study, we have evaluated a new collagen product,
R
, which is designed with a compact and spongy
Mucograft
layer in order to serve a barrier function for applications
in guided tissue regeneration. This represents the first in
vivo evaluation of this material. The tissue reaction to this
material is quite favorable, with minimal inflammation and
no multinucleated giant cells. The material persisted in the
tissue throughout the study, and while the spongy layer was
infiltrated early on, the compact layer remained impermeable
to invading cells for the first 30 days of the study. This
demonstrates promise by indicating preferential cell ingrowth
from the spongy surface. However, in order to truly realize
the promise of a biomaterial, it must be evaluated in the
clinical setting. For this reason, we have also included results
R
to treat
from a pilot clinical evaluation using Mucograft
gingival tissue recession in humans. Our results from this
study demonstrate great potential to reverse tissue recession
and promote more healthy gingival tissue. Additionally,
histological evidence corroborates our findings in the murine
model, with the compact layer preventing tissue ingrowth
as the scaffold becomes infiltrated from the spongy layer.
R
system demonstrates promise in
Overall, the Mucograft
murine in vivo studies and realizes this therapeutic promise in a
human application of the material. In terms of the translation
of biomaterials, much can be learned from suitable in vivo
experiments, and in this case these in vivo studies correctly
predicted the clinical outcome of the material, although the
latter must be confirmed by more extensive clinical studies.

References
[1] Dahlin C, Sennerby L, Lekholm U, Linde A and Nyman S
1989 Generation of new bone around titanium implants
using a membrane technique: an experimental study in
rabbits Int. J. Oral Maxillofac. Implants 4 1925
[2] Gottlow J 1993 Guided tissue regeneration using bioresorbable
and non-resorbable devices: initial healing and long-term
results J. Periodontol. 64 (11 Suppl) 115765
[3] Hammerle C H and Karring T 1998 Guided bone regeneration
at oral implant sites Periodontology 2000 17 15175
[4] Hammerle C H and Lang N P 2001 Single stage surgery
combining transmucosal implant placement with guided
bone regeneration and bioresorbable materials Clin. Oral
Implants Res. 12 918
[5] Karring T 1986 [The role of bone regeneration in post-surgical
periodontal wound healing] Orthod. Fr. 57 (Pt 2) 43542
[6] Schenk R K, Buser D, Hardwick W R and Dahlin C 1994
Healing pattern of bone regeneration in
membrane-protected defects: a histologic study in the
canine mandible Int. J. Oral Maxillofac. Implants 9 1329
[7] Weng D et al 2009 The effects of recombinant human
growth/differentiation factor-5 (rhGDF-5) on bone
regeneration around titanium dental implants in barrier
membrane-protected defects: a pilot study in the mandible
of beagle dogs Int. J. Oral Maxillofac. Implants 24 317
[8] Hoffmann O, Bartee B K, Beaumont C, Kasaj A, Deli G
and Zafiropoulos G G 2008 Alveolar bone preservation in
extraction sockets using non-resorbable dPTFE membranes:
a retrospective non-randomized study J. Periodontol.
79 135569
11

Biomed. Mater. 6 (2011) 015010

S Ghanaati et al

[9] Eickholz P, Kim T S, Holle R and Hausmann E 2001

Long-term results of guided tissue regeneration therapy
with non-resorbable and bioabsorbable barriers: I. Class II
furcations J. Periodontol. 72 3542
[10] Greenstein G and Caton J G 1993 Biodegradable barriers and
guided tissue regeneration Periodontology 2000 1 3645
[11] Bunyaratavej P and Wang H L 2001 Collagen membranes: a
review J. Periodontol. 72 21529
[12] Lang N P, Hammerle C H, Bragger U, Lehmann B and
Nyman S R 1994 Guided tissue regeneration in jawbone
defects prior to implant placement Clin. Oral Implants Res.
5 927
[13] Armstrong D G and Jude E B 2002 The role of matrix
metalloproteinases in wound healing J. Am. Podiatr. Med.
Assoc. 92 128
[14] Fuchs S et al 2009 Contribution of outgrowth endothelial cells
from human peripheral blood on in vivo vascularization of
bone tissue engineered constructs based on starch
polycaprolactone scaffolds Biomaterials 30 52634
[15] Fuchs S et al 2009 Dynamic processes involved in the
pre-vascularization of silk fibroin constructs for bone
regeneration using outgrowth endothelial cells Biomaterials
30 132938
[16] Minabe M et al 1989 Different cross-linked types of collagen
implanted in rat palatal gingiva J. Periodontol. 60 3543
[17] Pitaru S, Tal H, Soldinger M, Grosskopf A and Noff M 1988
Partial regeneration of periodontal tissues using collagen
barriers. Initial observations in the canine J. Periodontol.
59 3806
[18] Petite H, Rault I, Huc A, Menasche P and Herbage D 1990
Use of the acyl azide method for cross-linking collagen-rich
tissues such as pericardium J. Biomed. Mater. Res.
24 17987
[19] Goissis G, Marcantonio E Jr, Marcantonio R A, Lia R C,
Cancian D C and de Carvalho W M 1999 Biocompatibility
studies of anionic collagen membranes with different degree
of glutaraldehyde cross-linking Biomaterials 20 2734

[20] Rothamel D, Schwarz F, Sculean A, Herten M, Scherbaum W

and Becker J 2004 Biocompatibility of various collagen
membranes in cultures of human PDL fibroblasts and
human osteoblast-like cells Clin. Oral Implants Res.
15 4439
[21] Rothamel D, Schwarz F, Sager M, Herten M, Sculean A
and Becker J 2005 Biodegradation of differently
cross-linked collagen membranes: an experimental study in
the rat Clin. Oral Implants Res. 16 36978
[22] Schwarz F, Rothamel D, Herten M, Sager M and Becker J
2006 Angiogenesis pattern of native and cross-linked
collagen membranes: an immunohistochemical study in the
rat Clin. Oral Implants Res. 17 4039
[23] Ghanaati S et al 2010 Fine-tuning scaffolds for tissue
regeneration: effects of formic acid processing on tissue
reaction to silk fibroin J. Tissue Eng. Regen. Med. 4 46472
[24] Ghanaati S et al 2009 Dynamic in vivo biocompatibility of
angiogenic peptide amphiphile nanofibers Biomaterials
30 620212
[25] Zucchelli G and De Sanctis M 2000 Treatment of multiple
recession-type defects in patients with esthetic demands
J. Periodontol. 71 150614
[26] Jones J A et al 2004 Macrophage behavior on surface-modified
polyurethanes J. Biomater. Sci. Polym. Ed. 15 56784
[27] Keselowsky B G et al 2007 Role of plasma fibronectin in the
foreign body response to biomaterials Biomaterials
28 362631
[28] McNally A K, Jones J A, Macewan S R, Colton E
and Anderson J M 2008 Vitronectin is a critical protein
adhesion substrate for IL-4-induced foreign body
giant cell formation J. Biomed. Mater. Res.
A 86 53543
[29] Ghanaati S et al 2010 Histological and histomorphometrical
analysis of a silica matrix embedded nanocrystalline
hydroxyapatite bone substitute using the subcutaneous
implantation model in Wistar rats Biomed. Mater.
11 035005

12